Search

Abstract Bio-derived solvents such as ρ-cymene , d-limonene and α-pinene represent feasible alternatives to n-hexane for the extraction of vegetable oils. However, the large-scale utilization of these solvents is still limited mainly owing to their high boiling points and latent heats of vaporization. In this work, the performance of composite polydimethylsiloxane/polyacrylonitrile (PDMS/PAN) organic solvent nanofiltration membranes in the recovery of these solvents from their binary mixtures with canola oil is investigated. The sorption isotherms of the mixtures were first studied using free-standing PDMS films and the multicomponent Flory–Huggins model used to determine the resulting interaction parameters. The partial solvent uptake decreased with increasing oil concentration in the mixture. On the other hand, the partial oil uptake in the solvent mixture was higher than that of the pure oil which was attributed to the swelling effects induced by solvents. The effects of feed concentration (10–30 wt% oil), feed temperature (25–40 °C), transmembrane pressure (5–30 bar), and cross-flow velocity (18–52 cm s−1) on the membrane performance were then studied in a cross-flow membrane setup. Maxwell–Stefan formulations were combined with the ternary Flory–Huggins solubility model to successfully describe these flux data. Highlights • Sorption isotherms determined for 10 and 30 wt% canola oil in terpene mixtures into PDMS films. • Sorption data predicted by a ternary Flory–Huggins model. • Nanofiltration flux and rejection of these oil and terpene mixtures measured through PDMS films. • Flux data fitted to a Maxwell–Stefan ternary mixture diffusion model. [ABSTRACT FROM AUTHOR]

The removal of natural organic matter (NOM) from boreal lake water by a novel capillary nanofiltration (NF) membrane was predicted using a computational fluid dynamics (CFD) modelling approach. The 2-dimensional axis-symmetric model was based on a 48 m 3 /day NF pilot plant operating in cross-flow mode on water containing 8 mg/L total organic carbon (TOC) at fluxes ranging from 10 to 25 L/m 2 /h and velocities ranging from 0.25 to 1.0 m/s. A “mass jump” source code developed using the solution diffusion model was used to simulate water flux and variations in NOM content as a function of axial and radial position in the capillary fibres. The model was validated within 3% inaccuracy using pilot data for filtrate TOC and UV254 absorbance and longitudinal pressure drop. The model was subsequently used to compare the effect of module length and number of stages on the design performance of a 110,000 m 3 /day NF plant. Simulations indicated that 1.5 m long modules operated in a double pass configuration removed 33% more NOM compared with 3.0 m long modules in a single pass. Moreover, the use of 1.5 m modules in the full-scale plant configured in a 10:5:3:2 four stage array achieved greater NOM removal than a 10:5:3 three stage at the same plant water recovery (90.5%) using lower recycle rates and lower net energy consumption. The paper demonstrates that the combination of experimental and numerical methods can be an effective tool for the design of nanofiltration plants for enhanced NOM removal. [ABSTRACT FROM AUTHOR]

A preliminary study was carried out on the morphological design of tubular single-channel alumina membranes prepared by stereolithography, an additive manufacturing process. The geometry of the ring-patterned inner surface of membranes was optimized using computational fluid dynamics calculations and validated in microfiltration tests with aqueous suspensions of P. aeruginosa. Patterning of the inner surface of tubular membranes helped reduce cake formation at a higher value of the average crossflow velocity. The results highlight benefits of stereolithography-based approach to the morphological design of ceramic membranes. [Display omitted] • 3D-printed patterned tubular ceramic membranes are prepared by stereolithography. • CFD modeling guides membrane surface morphology design at the mesoscale. • Ring-patterning of the feed-facing surface helps alleviate membrane fouling. • Crossflow microfiltration tests with P. aeruginosa corroborate CFD predictions. • Additive manufacturing shows promise for morphological design of ceramic membranes. [ABSTRACT FROM AUTHOR]

Microfiltration is an attractive means for treating oily wastewater, especially when the size of the oil droplets are micrometer-sized since the conventional techniques become deficient. A systematic study on the critical flux of oil-in-water emulsion, which behaves differently from other colloidal foulants with regards to deformation, coalescence and splitting, has not been carried out to date. This was the goal of the current study, which employed the Direct Observation Through the Membrane (DOTM) technique to characterize the critical flux of oil-in-water emulsions of various concentrations, and at various cross-flow velocities (CFV) and salt concentrations. Five observations can be highlighted here. Firstly, the oil droplets with a mean droplet diameter of approximately 5 µm exhibited critical fluxes equal to or greater than latex particles of 10 µm. This is likely due to the twin effects of membrane oleophobicity promoting back-transport of the oil foulant from the membrane and the presence of a droplet size distribution with larger drops that can enhance the shear-induced diffusion of the average drops. Secondly, the critical flux values did not agree with the model that is valid for the size range the mean droplet diameter falls in, but instead agreed with the model adapted for smaller particulate foulants. Thirdly, the increase in the critical flux with CFV was more significant for the lower oil concentration. Fourthly, a striping phenomenon was observed at higher oil concentrations and lower CFV values. Striping was not observed for latex particles. Fifthly, the critical flux decreased with salt concentration. These findings highlight the unique fouling behavior of oil-in-water emulsions in microfiltration. [ABSTRACT FROM AUTHOR]

Critical and threshold flux concepts were recently developed to distinguish no fouling, slow fouling and rapid fouling regimes. Membrane fouling behavior is expected to vary with respect to the imposed flux relative to the critical and threshold flux values. However, crossflow fouling tests are often performed independent of critical and threshold flux determinations. In this study, constant flux fouling experiments were performed in connection with critical and threshold flux determination. Fouling behavior was examined in the context of critical and threshold flux. A poly(vinylidene fluoride) microfiltration membrane was challenged with various oil-in-water emulsions. The critical and threshold flux values were estimated using the flux-stepping technique. Constant flux crossflow fouling tests were performed at selected fluxes below and above the critical and threshold fluxes. Below the critical flux, mass transfer resistance remained constant at the clean membrane value. Above the critical flux but below the threshold flux, mass transfer resistance approached a steady state resistance, R B , which was determined from the linear regression of flux-stepping experiments. Above the threshold flux, a three-stage transmembrane pressure (TMP) was observed, consisting of: (1) an initial gradual increase, (2) a TMP jump stage, and (3) a pseudo-steady state. The pseudo-steady state TMP corresponded to the estimated critical pressure of the oil layer. [ABSTRACT FROM AUTHOR]

Membrane fouling is an endemic problem in ultrafiltration used for many applications. Recently water-hammer pulsing has been shown to reduce membrane fouling. Rapid solenoid-valve closure on the retentate or permeate side of a membrane generates a pressure front that propagates and reflects at the system boundaries at the speed-of-sound. However, the mechanism whereby this mitigation occurs is not well-understood. A predictive model is developed for mitigating fouling using pulsed solenoid-valve closure in the ultrafiltration of particulates. The model assumes the boundary-layer flow that develops after the first pressure-front reflection at the upstream boundary causes an enhanced shear stress on the fouled membrane. This increased shear stress typically occurs for less than 0.01 s during each solenoid-valve closure that occurs once every several seconds. However, it has a significant effect on removing the fouling deposits because it is orders-of-magnitude larger than the shear stress of the steady state crossflow velocity. This model correlates with a Coefficient-of-Determination of 0.997 the fractional flux increase for 44 experiments spanning a range of fluxes, foulant concentrations, pressures, and solenoid-valve-closure frequencies for an aqueous whey-protein feed. The fractional flux increase was 10.7%–124% and averaged 51.2%. However, the model indicates that considerably higher flux increases are possible. Predicted fractional permeation flux increase (J ′‒ J)/ J as a function of αf / J for mitigating membrane fouling in the ultrafiltration of an aqueous whey-protein feed using water hammer; J ′ and J are the permeation fluxes in the presence and absence of water hammer, respectively, f is the solenoid-valve closure frequency, and α is a semiempirical parameter. [Display omitted] • Pulsed solenoid-valve closure can increase UF flux in the presence of fouling by as much as 124%. • Mechanism proposed for mitigation of particulate membrane fouling in the presence of water hammer. • Boundary layer generated by pressure-front reflection causes high shear stress that mitigates fouling. • Predictive model advanced that incorporates effects of flux, valve-closure frequency, pressure, and concentration. • Model correlates 44 data with CoD of 0.997 for UF concentration of aqueous whey-protein feed. [ABSTRACT FROM AUTHOR]

The influence of cross-flow velocity and transmembrane pressure on the optimal back-shock frequency and normalized net flux during back-shocking has been studied using a semi-analytical mathematical model. The model uses the flux as a function of time without back-shocking together with knowledge of the streamlines and pathlines during the back-shock cycle to predict the optimal normalized net flux as a function of forward filtration time. The model was used to investigate three different transmembrane pressures and three different cross-flow velocities during a back-shock cycle. The net flux was found to increase under all operating conditions when using back-shocking. The greatest increase in normalized net flux was found at the highest cross-flow velocity and the highest transmembrane pressure, and corresponds to an increase of 37% compared to the steady-state flux. The highest cross-flow velocity and the highest transmembrane pressure gave the highest optimal back-shock frequency of 0.21 Hz. The optimal back-shock frequency was found to decrease with increasing pressure and decreasing cross-flow velocity. The model is easy to use in different applications as it is easy to measure flux during forward filtration without back-shocking. Good agreement was found between the semi-analytical model and a model based on computer fluid dynamics in predicting both the value of the optimal normalized net flux and the optimal back-shock frequency. [ABSTRACT FROM AUTHOR]

The influence of back-shock frequency on the net flux at different transmembrane pressures and cross-flow velocities has been studied. The mathematical model used in the investigation solves the continuity equation in a two-dimensional domain near the membrane surface. An osmotic pressure model is used to calculate the concentration at the membrane surface. The highest value of the normalized net flux was found at high transmembrane pressure and high cross-flow velocity. The normalized net flux at optimal conditions was 1.36 times higher than the flux without back-shocking. The optimal forward filtration time at the highest transmembrane pressure and the highest cross-flow velocity is approximately 5 s, corresponding to a frequency of 0.18 Hz. The optimal forward filtration time decreases as a function of both increasing transmembrane pressure and increasing cross-flow velocity. [ABSTRACT FROM AUTHOR]

A fundamental study of microfiltration membrane fouling by emulsified oil was conducted using a combination of real-time visualization, force balance on a droplet, and permeate flux analysis. The model 0.1% v/v hexadecane-in-water emulsions contained sodium dodecyl sulfate (0.1 mM, 0.4 mM, or 0.8 mM) to regulate interfacial tension. Direct Observation Through the Membrane tests with Anopore ( d pore =0.2 µm) and track-etch ( d pore =5 µm) membranes revealed three characteristic stages of membrane fouling: (1) droplet attachment and clustering, (2) droplet deformation, and (3) droplet coalescence. In qualitative agreement with visualization results, the force balance predicted that droplets ≲ 36–40 µm would remain pinned at d pore =5 µm pores while larger droplets would be swept off the surface by the crossflow drag. In a separate set of constant pressure crossflow filtration tests with track-etch membranes, the average oil rejection was ≥98% while the permeate flux decreased to a pseudo-steady-state ~10% of the initial value. The results indicate that membrane fouling by emulsified oil is controlled by droplet coalescence and crossflow shear: the transport of oil to the membrane surface by the permeate flow is balanced by the shear-induced removal of the droplets that coalesce to exceed a critical size. [ABSTRACT FROM AUTHOR]

Small devices of cross-flow nanofiltration (NF) may become a tool to concentrate small samples of valuable substances, if the concentration polarization can be minimized by designing new feed channels with bas-reliefs similar to the ones that have been developed for efficient passive micromixers. These bas-reliefs may have a complex 3D shape and the recourse to intensive CFD simulations is necessary to optimize the geometry. To speed up the simulations, a hybrid, semi empirical approach is used where the average and local mass transfer coefficients are first computed by CFD, using a dissolving wall instead of a semi-permeable membrane boundary condition. The computed mass transfer coefficients are then used to predict the local and average concentration polarization, using a mass transfer correction factor correlation. This computational approach was validated using detailed CFD simulations and experimental concentration distributions obtained by holographic interferometry. A NF ribbed-wall rectangular channel with 2 mm height was used, with potassium sulfate solutions, transmembrane pressures lower than 8 bar and Reynolds numbers between 1 and 125. Using this computational approach, local and average concentration polarization and permeate flux were predicted with an error less than 15%. With these findings, it is possible to speed up the simulation and design of small nanofiltration cross-flow devices. [ABSTRACT FROM AUTHOR]

The main drawback of membrane processes is biological membrane fouling which, in turn, leads to the membrane having an increased amount of transmembrane pressure and energy consumption. The fouling of microfiltration (MF) and ultrafiltration (UF) membranes caused by microalgal cells was investigated in this study. An MF membrane with a pore size of 0.20 µm (made of PVDF) and UF membranes with a molecular weight cutoff (MWCO) of 150, 50, and 30 kDa (made of PES, PESH and RC, respectively) were employed for the filtration experiments. An attenuated total reflection-Fourier infrared (ATR-FTIR) spectroscopy, a scanning electron microscope (SEM), and contact angle device were utilised in order to characterise the membranes fouled by microalgal cells. The results demonstrated that the MF membrane exhibited a faster decrease in flux when the normalised flux was reduced to as low as 0.15. An increase in the crossflow velocity (CFV), on the other hand, can dramatically decrease microalgal cells depositing on the surface of a membrane, thereby leading to higher flux. Moreover, microalgal cells caused both reversible and irreversible fouling for all membranes. In addition, the analysis of the different filtration resistances encountered in membrane filtration involving concentration polarisation, cake layer formation, and the clogging of pores was studied. The experimental results obtained in this study show that the cake resistance ( R c ) had the highest resistance for the UF membranes and that the concentration polarisation ( R cp ) was dominant in all crossflow velocities for MF membranes. [ABSTRACT FROM AUTHOR]

There is some controversy in the literature on the effects of combined fouling of organics and colloids on water flux. There are reports of more flux decline for combined fouling than that of fouling by individual foulants, which has been attributed to synergistic effects. However, some authors observed lower flux decline in combined fouling. In addition, some authors have reported on increased flux declines by calcium addition, whereas others have shown improved fluxes after calcium addition. The discrepancies observed in the literature are systematically investigated in more detail in this manuscript. The organic–colloid concentration ratio was varied in order to explain discrepancies in previously observed effects in combined fouling. Further, the probable contributing factor of colloid type and calcium concentration in combined fouling was also examined. Organic, colloidal and combined fouling was found to be aggravated in the presence of Ca 2+ due to organic–Ca 2+ complexation and the reduction in foulant–membrane repulsive interactions, but this only occurred up to certain Ca 2+ -concentrations. Above these concentrations, the opposite effect was observed. In addition, membrane–foulant as well as foulant–foulant affinity interactions were determined and related to initial and later fouling rates. A novel type of “sequential” fouling experiments were carried out for the first time, and yielded information on how colloids and organics interact when both are present in fouling solutions. It was concluded from these experiments that hindered back diffusion of organics by a layer of colloids was the main factor responsible for flux decline in combined fouling. [ABSTRACT FROM AUTHOR]

Patterned microfiltration membranes were fabricated via Phase Separation MicroFabrication (PSMF) using vapor-induced phase separation to avoid skin formation on the nonpatterned surface. A starting solution of 10% PES, 60% PEG400, 5% water and 25% NMP yielded symmetric membranes with similar pore size all throughout the cross section when 10 or 15 min of vapor exposure was used before further coagulation in water. On the other hand, 5 min of vapor exposure resulted in what may be called inverse asymmetric membranes, which had smaller pores on the patterned (mold) side and macrovoids throughout the cross-section. The fidelity in replicating the pattern was higher when the vapor exposure was shorter and surface area enhancements of 103% and 52% were obtained when vapor exposure times were 5 and 10 min, respectively. The phase inversion rates of the polymer solutions in liquid water, humid air and humid air followed by liquid water were examined by measuring the light transmission through the solutions in time. It was observed that water in the starting solution increased the overall phase inversion rate and decreased the delay time before phase separation, in consistence with the membrane morphologies obtained with the solutions with and without water under different coagulation conditions. The patterned and nonpatterned membranes were tested in constant flux crossflow filtration of yeast suspensions at different fluxes and it was seen that the fouling rate of nonpatterned membranes was higher than patterned membranes. However, the major contribution to the improved fouling behavior appears to be the increased surface area since when the fluxes were normalized by the patterned (actual) surface area for foulant deposition, the difference between the fouling rate of patterned and nonpatterned membranes was not significant. [ABSTRACT FROM AUTHOR]

Pressure-retarded osmosis (PRO) has the potential to convert salinity gradient energy to hydraulic energy via a semi-permeable membrane. The performance of an osmotic membrane is usually evaluated using a crossflow membrane test cell with symmetric channels. However, it is not possible to test a PRO membrane under conditions of large differences in hydraulic pressure because the membrane cannot be supported in the channel inlet and outlet with a cavity. In this study, we present a newly designed PRO crossflow test cell that features asymmetric channels on both sides of the membrane. This asymmetric design prevents membrane bulge and rupture in the channel inlet and outlet even at an applied hydraulic pressure difference greater than 54.5 bar. Excellent performance of this novel PRO cell was demonstrated using two different membranes at various operating pressures. To prevent membrane deformation at open spaces between spacer strands, a feed channel spacer should be selected carefully. Accordingly, the effects of tricot feed spacers with different voidages and thicknesses on PRO performance were also investigated. [ABSTRACT FROM AUTHOR]

In this investigation, a simplified pressure retarded osmosis (PRO) model incorporating the detrimental effects of internal polarization concentration (ICP), external polarization concentration (ECP) and reverse salt permeation (RSP) is proposed and verified using published data. The results demonstrate the accuracy of the model to address decreased water flux and power density due to the performance limiting effects. Based on the model, the discharge behaviour of a PRO process is reported with respect to different applied pressures on the draw solution and two flow schemes, co-current and counter-current flows. In the co-current flow PRO process, from the flow profiles in the draw and feed channels, it is found that the adverse effects on the process dynamics, such as water flux and power density, and required membrane area, can be regarded as a further retardation by applying ‘an extra applied pressure’ on the draw solution. In addition, the capacity of extractable energy of the full scale PRO discharge is significantly reduced due to the ICP, ECP and RSP effects. Furthermore, the termination conditions of the PRO discharge and its effects on the dynamics of the PRO process are also investigated in the case of the counter-current flow PRO process. [ABSTRACT FROM AUTHOR]

Previous work has suggested that membrane patterning is a promising approach to fouling mitigation. In this systematic study, we describe performance metrics for the cross-flow filtration of colloidal suspensions through ultrafiltration membranes with lithographically patterned surfaces. The effects of particle size, cross-flow velocity, pattern size, and flow configuration on membrane performance are specifically considered. Using an unpatterned membrane as a reference, results show that the presence of surface patterns increases the critical flux associated with filtration of colloidal feed solutions. For the patterned membranes, the critical flux increases with particle size, cross-flow velocity, the angle between the feed flow and the pattern lines, and the pattern height. These experimental findings can be correlated with a back-transport mechanism such as shear-induced diffusion, which ultimately increases the threshold permeation flux associated with the onset of particle deposition. [ABSTRACT FROM AUTHOR]

Biofilm development in a spacer-filled reverse osmosis membrane channel can influence both trans-membrane pressure (TMP) and channel pressure drop (ΔPCH). While current pretreatment methods are unable to completely tackle the biofouling problem, more insights are required to provide strategies to minimize the problem. This study examined the role of operating parameters (i.e. flux and crossflow velocity) to minimize biofouling in RO processes. The experiments were conducted with a lab-scale high pressure flat sheet RO reactor where changes in pressure drop along the channel and across the membrane were measured. The impact of biofouling was measured at constant fluxes, where the TMP rise and ΔPCH rise and the biofoulant was quantified as biovolumes of live and dead bacteria on autopsied membrane and spacer samples by confocal laser scanning microscopy (CLSM). The results show that TMP rise increased exponentially with increasing flux, and decreased with increasing crossflow velocity. The channel pressure drop, ΔPCH, increased when either flux or crossflow velocity was increased, and was more dependent on crossflow. The biofoulant volume on the membrane increased with flux and was less dependent on crossflow. The biofoulant associated with the spacer was much less than on the membrane and relatively insensitive to flux or crossflow velocity. The TMP rise could be correlated with the estimated concentration of nutrient at the membrane surface, Cw,N, highlighting the combined roles of flux and crossflow velocity in solute concentration polarization. Previous TMP rise data could also be correlated to the estimated Cw,N values. This observation suggests a biofouling mitigation strategy by controlling both incoming nutrient concentration and operating conditions (flux and crossflow). [ABSTRACT FROM AUTHOR]

Abstract: The behavior of an oil droplet pinned at the entrance of a micropore and subject to crossflow-induced shear is investigated numerically by solving the Navier–Stokes equation. We found that in the absence of crossflow, the critical transmembrane pressure required to force the droplet into the pore is in excellent agreement with a theoretical prediction based on the Young–Laplace equation. With increasing shear rate, the critical pressure of permeation increases, and at sufficiently high shear rates the oil droplet breaks up into two segments. The results of numerical simulations indicate that droplet breakup at the pore entrance is facilitated at lower values of the surface tension coefficient, higher oil-to-water viscosity ratio and larger droplet size but is insensitive to the value of the contact angle. Using simple force and torque balance arguments, an estimate for the increase in critical pressure due to crossflow and the breakup capillary number is obtained and validated for different viscosity ratios, surface tension coefficients, contact angles, and drop-to-pore size ratios. [Copyright &y& Elsevier]

Abstract: We introduce a new type of membrane spacer design, based on the geometry of static mixers, for application in submerged membrane filtration. In particular we demonstrate the potential of the spacers to reduce air sparging and energy consumption. New operational modes are proposed and discussed critically. Critical flux measurements are performed with and without spacer at various process parameters such as spacer/channel height, air sparging rate and bubble size. The most relevant energy reducing parameters of submerged systems are determined with the Design of Experiments (DoE) methodology. Using the new spacers, the critical flux can be increased by 100%. Furthermore, fouling monitored by the trans-membrane pressure change can be significantly reduced by a factor of 7.5 when combined with elevated cross-flow velocities. [Copyright &y& Elsevier]

Abstract: Fouling, more precisely biofouling, remains one of the main obstacles for a wider application of membrane technology. One of the most significant problems limiting membrane fouling control is the inadequate understanding of the basic mechanisms governing adsorption/deposition of foulants in and on membranes. A detailed knowledge with regard to fouling characteristics and development would be clearly advanced if a reliable and repeatable, non-invasive study tool would be available. For this purpose, a direct visual observation system, containing multiple separated cross-flow filtration units, coupled with either an optical or a confocal laser scanning microscope was developed. This approach allows the online, non-destructive monitoring of the fouling process while filtering and the high number of parallel filtrations makes it possible to either analyze different membranes, to apply statistical evaluations, run different feed solutions or to sample and perform membrane autopsies at different time points. This technique allows for the investigation of microbial ecology with respect to microbe–membrane attachment, site-specific associations of micro-organisms, deposition of foulants on the membrane surface and spatial organization of microbial communities on membrane surfaces through the use of fluorescent microscopy techniques. This manuscript illustrates the reproducibility of the developed research tool with respect to membrane flux decline and fouling characterization. Two membrane types, a polyvinylidene fluoride microfiltration (under filtering conditions) and a polyamide reverse osmosis membrane (in non-permeating mode), with respectively a bacterial foulant, Pseudomonas aeruginosa and tap water, were chosen as model systems. [Copyright &y& Elsevier]

In membrane filtration, the pore size of the membrane determines the size of ‘particles’ that should be rejected, leading to accumulation of particles on the membrane surface and changed particle retention in time. A process without accumulation and thereby constant retention as function of time would be well suited for fractionation of components close in size. In this research, emulsions consisting of small droplets (∼2.0µm) and large droplets (∼5.5µm), with total concentrations between 10% and 47%, were fractionated. The cross-flow module consisted of a closed channel to allow particles to migrate, followed by a membrane area with 20µm pores where emulsion fractions could be removed. Under appropriate process conditions, the permeate consisted of only small droplets, at concentrations higher than in the original emulsion, leading to very high selectivities. Especially at high concentrations, known to cause severe fouling in regular membrane filtration, these effects were occurring as a result of shear-induced diffusion of the droplets. If only small particles are targeted, the module can be operated at fluxes of 40L/(m2/h); if fractionation is targeted the fluxes can be considerably higher. These fluxes are comparable to current operational fluxes, but here cross-flow velocity and trans-membrane pressure are much lower (corresponding to fluxes of 1–4m3/(m2/h/bar)) with stable retention and flux as function of time. [ABSTRACT FROM AUTHOR]

Abstract: Analytical solutions of the Nernst–Planck, Poisson and continuity equations for a membrane undergoing osmosis in a cross-flow system reveal that the flow of alternating ionic charge induced in the membrane during impedance measurements is actively assisted by the flow of water. The solutions yield expressions for the impedance of the membrane that characterize the dielectric, conduction and geometrical properties (Maxwell–Wagner model) as well as the actively assisted contributions that are functions of the velocity of water and the electrical field values at the membrane boundaries. The consistency of this theory with experimental results reported in literature provides a basis for interpreting in situ impedance characterizations of membrane systems, and in particular the viability of the membranes and the extent of fouling during reverse osmosis. [Copyright &y& Elsevier]

Abstract: Cross-flow microfiltration of fermented mash containing native corn starch granules from a simultaneous saccharification and fermentation process, was investigated with respect to the process performance. Influence of transmembrane pressure and cross-flow velocity on normalized permeate flux rate decline and membrane fouling were examined. The microfiltration system was equipped with a tubular ceramic membrane with 0.45μm pore size. A combined pore-blockage–cake-filtration model was successfully used to determine the process performance and identify principal fouling mechanisms. Experimental results revealed that the fouling phenomenon taking place during microfiltration of the fermented mash follows a pattern of initial complete pore blockage and subsequent cake formation. As a result, initial rapid flux decline is followed by gradual flux decline. Fouled membranes were characterized using scanning electron microscopy (SEM). The membrane surface characteristics examined by SEM images evidenced formation of dense deposit on the upper membrane surface. Rinsing the fouled membrane with water followed by investigation of the permeate flux rate indicated occurrence of increased fouling, when microfiltration was carried out at high transmembrane pressure. The highest normalized flux decline was observed for microfiltration operated at the highest transmembrane pressure (1.4bar) and the lowest cross flow velocity (1ms−1). It is recommended to use high cross-flow velocity (4.55ms−1) and low transmembrane pressure (0.35bar) to minimize membrane fouling, triggered by components of the fermented mash. [Copyright &y& Elsevier]

Abstract: Intensification of microfiltration has been accomplished using motionless mixers consisting of a series of pairs of semielliptical blades as turbulence promoters. Blade mixers of two aspect ratios 2.5 and 1.3 were tested in the microfiltration of milk (0.1μm membrane). The permeate fluxes are substantially increased by application of blade mixers due to a reduction of both reversible and irreversible fouling. The highest flux improvements of 500–650% for the same cross-flow rate (relative to the conventional operation) were obtained by application of the blade mixer of aspect ratio 1.3.When compared for the same hydraulic dissipated power mixer of aspect ratio 2.5 proved to be slightly more efficient because it causes the lower pressure drop. Despite to the increased pressure drop, the energy savings obtained by application of blade mixers are considerable compared to the conventional operation and to the operation using some other mixers. In the membrane fitted with a blade mixer the flow field changes in a manner which afford the intensive disruption of boundary layer, scouring and removal of fouling forming material. The flow field is characterized by the high cross-flow velocities at membrane wall, alternation of stream line path and swirling motion. Therefore, the application of blade mixers affords an operation under the several times lower initial cross-flow velocity. [Copyright &y& Elsevier]

Abstract: Critical fouling conditions were studied during wine cross-flow microfiltration using a multichannel ceramic membrane (0.2μm). The aim was to determine critical operating conditions in order to limit fouling caused by wine colloids (tannins, pectin and mannoproteins) and enhance process performances. The method used is a square wave filtration based on the determination of the reversibility and irreversibility of fouling. Filtrations were performed with filtered red wine (FW) added with different concentrations of colloids. Considering FW, critical flux for irreversibility was beyond the studied range of pressure (≥1.4×10−4 m/s). No clear critical flux could be determined for any of the tested molecules in the studied range of pressure. On the other hand, an upper limit of fluxes range has been identified (below which critical flux could be found). Irreversible fouling always takes place from the beginning of the filtrations and even at low pressures. For FW containing 0.2g/l mannoprotein and 0.5g/l pectin, a loss of average fluxes is observed beyond a given limit of transmembrane pressure. This fact was attributed to the compaction of a gel layer. Finally, a criterion (R if/R m ≤1) has been suggested to determine the so-called “threshold flux” below it, fouling remains acceptable. [Copyright &y& Elsevier]

Abstract: The cross-flow microfiltration applied to wine filtration has become a legitimate alternative to conventional filtration processes. However, membrane fouling which affects the operating costs and the plant maintenance, limits the widespread application of this technique. The aim of this review is to provide a better understanding of the development of the cross-flow microfiltration in wine industry, as well as the complexity of wine composition and its consequences on membrane fouling. This review covers also the impact of the operating conditions and the membrane characteristics on fouling mechanisms. Strategies to limit fouling as well as the latest innovations and commercial proposal are discussed in this paper. [Copyright &y& Elsevier]

Abstract: In the dairy industry, whey protein is concentrated by means of cross-flow ultrafiltration (UF). In cross-flow UF, it is considered that the surface part of the fouling layer can be removed by the effect of the shear stress exerted by the feed flow. As a result, the flux decline is suppressed and a steady state flux eventuates. However, the means to express the relationship between the feed flow and the permeate process has not been come to definite conclusion. In this study, the compressive yield stress model for the permeation through a fouling layer on a membrane in dead-end UF was applied to cross-flow UF considering the removal rate of the fouling layer. The time course of permeate flux estimated based on the model. In the estimation, the concentration at the top of the fouling layer must be specified and the concentration distribution in the layer should be calculated. The rheology of a whey protein concentrate (WPC) suspension, which is used as a model fluid, was investigated. The concentration at the top of the fouling layer was determined based on the relation between the shear stress exerted by the feed flow and the rheology investigated. The validity of the model was discussed by comparing the results of estimation by the model and experiments. [Copyright &y& Elsevier]

Abstract: Electric field-assisted cross-flow ultrafiltration of mosambi (Citrus sinensis (L.) Osbeck) juice has been carried out using a 50kDa molecular weight cut-off membrane. Electric field is applied in both constant as well as pulse mode. The effect of process parameters, e.g., electric field, pulse ratio, cross-flow velocity and transmembrane pressure on the permeate flux and energy consumption are studied. Application of dc electric field has resulted in significant improvement of permeate flux. Application of pulsed electric field is found to be more beneficial in terms of energy consumption, and equally effective in terms of flux augmentation as compared to constant field. An already developed resistance-in-series model based on gel layer theory is used incorporating the effect of constant as well as pulsed electric field and numerically solved to quantify the flux decline and gel layer deposition. The predictions from the numerical solution of the model are successfully compared with experimental results under various operating conditions. [Copyright &y& Elsevier]

Abstract: The purpose of the study was to investigate the change in membrane resistance in relation to operating pressure for clean RO and NF membranes during filtration of a bioreactor effluent. Experiments were conducted using a SEPA CFII membrane element cell and effluent from a bioreactor with high-ionic content. Flux–pressure profiles of two RO and one NF membranes were analyzed and compared with those for filtration of deionized water. The flux–pressure profile increased linearly with increasing pressure for all three membranes during filtration of deionized water. However, during filtration of the bioreactor effluent, the rate of increase in flux decreased after the pressure reached a critical level with both RO membranes. The changes in flux and membrane resistance were analyzed in relation to the operating pressure which was correlated to transmembrane pressure using logarithmic and second order polynomial correlations. Experimental results showed that a second order polynomial function described the flux–pressure profile more adequately than the linear correlation. The use of the flux ratio (i.e., flux with bioreactor effluent to flux with deionized water) allowed a basis for comparison of change in membrane resistance with transmembrane pressure for different membranes. The critical pressure which resulted in the minimum membrane resistance was determined by the change in the membrane resistance expressed in dimensionless form as the ratio of the observed membrane resistance with bioreactor effluent and deionized water. [Copyright &y& Elsevier]

A new model based on the removal mechanism of loose/compact cake and the re-deposition mechanism of foulants was proposed to predict the instantaneous fouling resistance (R f (t)) in the rinsing process of 0.1 μm PAN membrane fouled with activated sludge. The variation trend of R f (t) was also analyzed by the removal rate of loose/compact cake (r r-loose / r r-compact) and the re-deposition rate of foulants (r re-deposition). Moreover, the model was validated by using other foulants (yeast, actual activated sludge, HA and SA) and membranes (0.1 μm PES and PVDF membranes). The results showed that the model predictions had good agreements with experimental data at different cross-flow velocities (1.0, 1.5, 2.0 and 2.5 m/s) or temperatures (25, 30, 35 and 40 °C) (R 2 >0.9900). Besides, the variation trend of R f (t) could be divided into three stages: (i) R f (t) quickly decreased and the removal of loose cake was almost completed at rinsing time of t 1 (first transition point) (r r-loose = 0); (ii) R f (t) slowly decreased and R f (t) reached to minimum (r r-compact = r re-deposition) at t 2 (second transition point); (iii) R f (t) slightly increased after t 2 (r r -compact (t)< r re -deposition (t)). This model would provide an accurate method to analyze the variation trend of R f (t) in the rinsing process and a theoretical basis for the optimization of rinsing conditions. [Display omitted] • A rinsing model was proposed to predict instantaneous fouling resistance (R f (t)). • The effect of rinsing mechanism on the variation trend of R f (t) were was studied. • Two transition points on the variation curve of R f (t) were determined by the model. • The removal of loose cake depended mainly on the cross-flow velocity. • The increase of temperature mainly promoted the removal of compact cake. [ABSTRACT FROM AUTHOR]

Fouling continues to be a limiting factor in membrane processes. Pulsating feed flows have the potential to decrease fouling while maintaining a constant pressure drop over the module. This study quantifies the impact of pulsating flows on particle deposition in a forward osmosis system with spacers at a mean crossflow velocity of 14 cm s−1 and permeate water flux of 18 L m−2 h−1. Carboxylate modified polystyrene beads were tracked by fluorescence microscopy in situ and real-time during the first 4 h of deposition with feed flow pulsations up to a frequency of 10 Hz, while experiments were conducted with a constant crossflow velocity as a reference. Imaging at 30 s intervals revealed deposition rates of up to 2.6 times higher at steady state, as compared to flow pulsations of 10 Hz. Consequentially, average deposition was reduced by a factor of 3.5 from steady state (109 ± 67 beads mm−2) to pulsating feed flow (31 ± 17 beads mm−2) after 4 h of deposition. The results of this research provide new fundamental insights related to the constraints and advantages of using pulsating flows for fouling mitigation in forward osmosis systems. [Display omitted] • In-situ and real-time investigation of cake layer formation in forward osmosis. • Quantification of particle deposition with fluorescence microscopy. • Pulsating flows are a promising approach for fouling mitigation in forward osmosis. • A higher pulsation frequency and amplitude leads to less deposition on the membrane. • Pulsation parameters have a significant influence on the spatial distribution of deposition. [ABSTRACT FROM AUTHOR]

The aim of this study was to understand the limiting flux in cross-flow microfiltration of a colloidal suspension in terms of the balance of hydrodynamic forces on a particle. Numerical simulations were carried out to estimate the lift force resulting from the cross-flow, drag force resulting from the convective flux, and net force. Assuming that the limiting flux is the flux when the net force on a particle is zero, we analyzed the simulation data to determine the correlation of two particle Reynolds numbers: R e p _ J l i m i t , calculated from the limiting flux, and R e p _ u δ , calculated from the cross-flow rate. The results indicate that R e p _ J l i m i t is proportional to the square of R e p _ u δ when a particle is at a dimensionless distance (distance divided by diameter) of less than 6.7 from the cake layer, in spite of the experimental fact that R e p _ J l i m i t is proportional to the 1.5th power of R e p _ u δ . In contrast, R e p _ J l i m i t is proportional to the 1.5th power of R e p _ u δ when a particle is at a dimensionless distance of 48–60. This means that the force balance of a particle at such a position, not at the membrane surface, determines the limiting flux in cross-flow microfiltration of a colloidal suspension. Image 1 • Hydrodynamic force on a particle in cross-flow microfiltration is simulated. • We focus on drag force induced by flux and lift force induced by cross-flow. • The balance of forces on a particle in a viscous sublayer is discussed. • The relationship between the force balance and limiting flux is discussed. [ABSTRACT FROM AUTHOR]

A self-cleaning surface can effectively solve the accumulation problem of irreversible contamination of ultrafiltration membranes. In this work, a (polyvinylidene fluoride)-g-(polyacrylic acid) (PVDF-g-PAA) ultrafiltration membrane, hereafter denoted as FAS, was prepared by a one-pot method. A superhydrophilic self-cleaning ultrafiltration membrane, hereafter denoted as FASF, was then obtained based on the strong complexation of Fe3+ with the abundant carboxyl groups at the FAS surface, which then mineralized Fe3+ into β-FeOOH nanoparticles and deposited them onto the FAS surface. This deposition was confirmed by Fourier transform infrared spectroscopy, thermogravimetric analysis, X-ray diffraction, and X-ray photoelectron spectroscopy. Bovine serum albumin (BSA) cross-flow filtration experiments confirmed the high ultrafiltration performance of FAS and FASF. FASF exhibited highly efficient degradation of methylene blue (MB) solution under visible light photo-Fenton conditions. The optimum amount of acrylic acid was 6 wt% in the one-pot reaction system, and the alkaline coagulation bath enhanced the wettability of FAS. The β-FeOOH nanoparticle content of the surface-mineralized FASF tended to be saturated after 24 h. β-FeOOH can significantly enhance the hydrophilicity and wettability of FASF and further enhance the antifouling performance of BSA. After four cycles of cleaning operation under visible light photo-Fenton conditions, the flux recovery rate of BSA-contaminated FASF reached 96.2%, and the rejection rate remained stable. FASF had superhydrophilicity, high antifouling, and fast photocatalytic self-cleaning performances, indicating good application prospects in organic wastewater treatment. Image 1 • A superhydrophilic self-cleaning ultrafiltration membrane (FASF) was prepared. • FASF had visible light photo-Fenton reaction activity. • FASF was obtained by Fe3+ mineralization on the surface of PVDF-g-PAA membrane (FAS). • FAS was prepared by graft copolymerization of AA and PVDF through a one-pot method. • FASF had stable antifouling and visible light photo-Fenton self-cleaning performances. [ABSTRACT FROM AUTHOR]

Membrane fouling is a common phenomenon in anaerobic membrane bioreactor (AnMBR), deteriorating membrane performance and increasing the operation costs of AnMBR. In this study, β-phase polyvinylidene fluoride (PVDF) flat sheet membranes were successfully prepared by first mixing PVDF powder with ionic liquid [BMIM]PF 6 and subsequently casting them using thermally induced phase separation (TIPS), followed by high voltage polarization treatment to endow the PVDF membranes with piezoelectric properties. For the first time in AnMBR, we reported in this study on the piezoelectric PVDF membrane that generates in-situ vibrations to reduce membrane fouling by applying a given frequency and alternating voltage (AV). The results suggested that the steady-state water flux through the piezoelectric membrane in AnMBR operation increased along with the AV, enhancing it by up to 72.6% (at 20 V) compared to that of the membrane without applying AV. In addition, high water flux was obtained at the optimal operating frequency of 10 kHz when the cross-flow mode was taken into consideration in AnMBR operation in this study, although the resonance frequency with maximum amplitude of the membrane was 600 kHz. These results indicate that the piezoelectric β-PVDF membrane is effective at minimizing or alleviating the membrane fouling when the electric field simultaneously applied during AnMBR operation. During the 30-day operation of AnMBR with the β-PVDF membrane with poling, the COD removal rate was maintained at over 90%, even if the applied frequencies and voltages were changed during the operation. Image 1 • Piezoelectric β-PVDF (P-b-PVDF) membrane was made by blending PVDF with ionic liquid. • P-β-PVDF membrane in AnMBR with poling showed higher water flux and less fouling. • 30-day operation of P-β-PVDF membranes in AnMBR exhibited over 90% COD removal. [ABSTRACT FROM AUTHOR]

Deposition and flow paths of particles (live and inert) often control the first stages of "cake" (fouling) formation in membrane systems. Developments in crossflow cell design such as three-dimensional (3D) printing and state of the art imaging techniques enable real-time and in situ tracking of initial cake formation. In this work, high-resolution large-field fluorescent microscopy was used to concomitantly study and compare the deposition and flow paths of inert beads and Bacillus subtilis as biofilm forming bacteria. The membrane surface was imaged continuously (30 s intervals) during the initial stages of cake formation (< 4 h) in a 3D printed forward osmosis (FO) cell containing diamond-shaped spacers. Deposition patterns were analyzed using particle detection and counting for quantitative comparison. Flow paths and velocities (< 17 μm s−1) of beads were similar to those of B. subtilis , thus providing a reliable approximation for bacteria passing through the feed channel. However, spatiotemporal deposition of beads and bacteria were markedly different: Final bacteria deposition was 20 times lower than that of beads when normalized to the initial foulant concentration in the feed solution, although beads had a linear deposition increase, while bacteria deposition rose exponentially. Additionally, deposition of beads was homogeneously distributed within the spacer element compared to bacteria, which were mostly captured around the spacer filaments obstructing the flow. Our results provide a novel approach to quantify cake formation in membrane systems as well as new insights on the impact of inert and living particles on the deposition patterns and flow paths during the first stages of cake formation. These insights could be applicable to design new membrane surfaces and spacer shapes that minimize and delay fouling development. Image 1 • 3D printed FO crossflow cell with spacer for in situ and real-time imaging. • Fluorescence microscopy for concomitant quantification of deposition and flow paths. • Time and space-resolved observation of deposition during initial cake formation. • Subtly different flow paths of beads and Bacillus subtilis at the membrane surface. • Significantly different deposition in time and space between beads and B. subtilis. [ABSTRACT FROM AUTHOR]

Many studies have demonstrated a close relationship between fouling and fluid dynamics to control fouling formation. However, few have been carried out to validate the relationship in practice, owing to the lack of techniques to evaluate fouling with hydrodynamics. We investigated the organic fouling of humic acid (HA) and tannic acid (TA) at three crossflow velocities using a fluorescence-labelled organism bioimaging instrument (FOBI). Higher crossflow velocity generally enhanced filtration performance; however, the type of foulants induced different fouling formation patterns on the membrane. The monitoring system and computational model revealed that HA fouling had strong (R2 = 0.82) correlation with the shear rate (obtained from a velocity of 10.42 cm/s), while TA had weak correlation (R2 = 0.13). HA fouling showed a high correlation with velocity profile at 10 cm/s, but TA fouling occurred in the over 20 cm/s velocity zone. This is because TA had relatively high adhesive force acting to the membrane, resulting in irreversible fouling that was hardly mitigated by crossflow control. Therefore, this study demonstrated that fouling development was predominantly controlled by hydrodynamic conditions, but the actual fouling distribution could be influenced by the physicochemical properties of dissolved organic matters. • Real-time monitoring was performed to evaluate the organic fouling development. • High crossflow velocity induced less fouling development and a high water flux. • Humic acid fouling had similarity with velocity distribution at high velocity. • Less correlation with flow was observed in tannic acid fouling distribution. [ABSTRACT FROM AUTHOR]