A three mechanism model to describe fouling of microfiltration membranes

Mathematical modeling of flux decline during filtration plays an important role in both sizing membrane systems and in the understanding of membrane fouling. Protein fouling is traditionally modeled using one of three classical fouling mechanisms: pore blockage, pore constriction or cake filtration. Here, we have developed a mathematical model to describe flux decline behavior during microfiltration accounting for all three classical fouling mechanisms. Pore constriction was assumed to first reduce the size of internal pores. Pore blockage then occurs at the top of the membrane, preventing further fouling to the interior structure. Finally the foulants at the top of the membrane form a cake, which controls the late stages of the filtration. The model prediction shows excellent agreement with experimental data for 0.25 μm polystyrene microspheres filtered through 0.22 μm Isopore membranes (where pore constriction is expected to be minimal) as well as non-aggregated bovine serum albumin solution through hydrophobic Durapore membranes (where pore constriction is expected to dominate). The effects of different fouling mechanisms on the flux decline were characterized by the ratio of characteristic fouling times of the different mechanisms. In this way the model can provide additional insights into the relative importance of different fouling mechanisms as compared to an analysis by a single mechanism model or by derivative plots, and it can be used to provide important insights into the flux decline characteristics.     

Effects of nanoparticles on the ultrafiltration of surface water

The effects of nanoparticles on the fouling behavior of UF membranes were investigated by filtering river water containing natural organic matter (NOM). Self-dispersible carbon black (70–200 nm) was employed to model nanoparticles in natural water. The presence of nanoparticles transformed the mode of initial fouling from internal pore adsorption of NOM to intermediate pore blocking, which caused a significant flux reduction. The use of powdered activated carbon to adsorb organic micromolecules reduced internal pore fouling, but this effect on initial fouling mode did not much mitigate the overall flux decline. As filtration proceeded, cake filtration became the dominant fouling mode. The resistance-in-series model revealed that boundary-layer resistance contributed significantly to increased filtration resistance in the filtration of river water. The nanoparticles nullified boundary-layer resistance plausibly by removing organic macromolecules from river water, but aggravated cake resistance, which required chemical cleaning. Addition of calcium significantly increased the aggregate size of nanoparticles from 0.18–0.35 μm to 3.4 μm, and thus reduced pore blocking and total cake resistance.     

A Combined Pore Blockage and Cake Filtration Model for Protein Fouling during Microfiltration

Previous studies of protein fouling during microfiltration have shown significant discrepancies between filtrate flux data and predictions of the classical pore blockage, pore constriction, and cake filtration models. A new mathematical model was developed for the filtrate flux which accounts for initial fouling due to pore blockage and subsequent fouling due to the growth of a protein cake or deposit over these initially blocked regions. The model explicitly accounts for the inhomogeneity in the cake layer thickness over different regions of the membrane arising from the time-dependent blockage of the pore surface. The model was shown to be in excellent agreement with experimental data obtained during the stirred cell filtration of bovine serum albumin solutions through polycarbonate track-etched microfiltration membranes over the entire course of the filtration. The model provides a smooth transition from the pore blockage to cake filtration regimes, eliminating the need to use different mathematical formulations to describe these two phenomena. In addition, the model provides the first quantitative explanation for some of the unusual observations reported previously in investigations of protein microfiltration. The results provide important insights into the underlying mechanisms of protein fouling during microfiltration. Copyright 2000 Academic Press.     

In situ monitoring techniques for concentration polarization and fouling phenomena in membrane filtration

Membrane fouling and subsequent permeate flux decline are inevitably associated with pressure-driven membrane processes. Despite the myriad of studies on membrane fouling and related phenomena--concentration polarization, cake formation and pore plugging--the fundamental mechanisms and processes involved are still not fully understood. A key to breakthroughs in understanding of fouling phenomena is the development of novel, non-invasive, in situ quantification of physico-chemical processes occurring during membrane filtration. State-of-the-art in situ monitoring techniques for concentration polarization, cake formation and fouling phenomena in pressure-driven membrane filtration are critically reviewed in this paper. The review addresses the physical principles and applications of the techniques as well as their strengths and deficiencies. Emphasis is given to techniques relevant to fouling phenomena where particles and solutes accumulate on the membrane surface such that pore plugging is negligible. The relevance of the techniques to specific processes and mechanisms involved in membrane fouling is also elaborated and discussed.     

Modeling of concentration polarization and depolarization with high-frequency backpulsing

Rapid backpulsing to reduce membrane fouling during crossflow microfiltration and ultrafiltration is studied by solving the convection-diffusion equation for concentration polarization and depolarization during cyclic operation with transmembrane pressure reversal. For a fixed duration of reverse filtration, there is a critical duration of forward filtration which must not be exceeded if the formation of a cake or gel layer on the membrane surface is to be avoided. The theory also predicts an optimum duration of forward filtration which maximizes the net flux, since backpulsing at too high of frequency does not allow for adequate permeate collection during forward filtration relative to that lost during reverse filtration, whereas backpulsing at too low of frequency results in significant flux decline due to cake or gel buildup during each period of forward filtration. In general, short backpulse durations, low feed concentrations, high shear rates, and high forward transmembrane pressures give the highest net fluxes, whereas the magnitude of the reverse transmembrane pressure has a relatively small effect.Rapid backpulsing experiments with yeast suspended in deionized water performed with a flat-sheet crossflow microfiltration module and cellulose acetate membranes with 0.07 μm average pore diameter. The optimum forward filtration times were found to be 1.5, 3, and 5 s, respectively, for backpulse durations of 0.1, 0.2, and 0.3 s. Both theory and experiment gave net fluxes with backpulsing of about 85% of the clean membrane flux (0.022 cm/s = 790 l/m² h), whereas the long-term flux in the absence of backpulsing is an order-of-magnitude lower (0.0026 cm/s = 94 l/m² h).     

Experimental study of ultrafiltration membrane fouling by sodium alginate and flux recovery by backwashing

Membrane fouling is the major limitation for a broader application of membrane technology. One of the main causes of membrane fouling in advanced wastewater reclamation and in membrane bioreactors (MBR) are the extracellular polymeric substances (EPS). Among the main constituents in EPS, polysaccharides are the most ubiquitous. This study aims at a better understanding of the fouling mechanisms of EPS and the efficiency of backwashing technique, which is applied in practice to restore membrane flux. For that purpose, the evolution of fouling by sodium alginate, a microbial polysaccharide, is studied in ultrafiltration. Fouling experiments are carried out in a single fiber apparatus, aiming at identifying the significance of distinct fouling mechanisms and their degree of reversibility by backwashing. An important parameter considered in the study is the concentration of calcium ions, which promote sodium alginate aggregation and influence the rate of flux decline, the reversibility of fouling and rejection. A rapid irreversible fouling takes place due to internal pore constriction, at the beginning of filtration, followed by cake development on the membrane surface. With increased calcium addition, cake development becomes the dominant mechanism throughout the filtration step. Furthermore, fouling reversibility is increased with the increase of calcium concentration. A unique behavior of sodium alginate solution in the absence of calcium is also noted, i.e. the formation of a labile layer on the membrane surface, which is affected by the small cross-flow that exists inside hollow fibers, even in the nominally dead-end mode of operation.     

Analysis of humic acid fouling during microfiltration using a pore blockage–cake filtration model

Fouling by natural organic matter, such as humic substances, is a major factor limiting the use of microfiltration for water purification. The objective of this study was to develop a fundamental understanding of the underlying mechanisms governing humic acid fouling during microfiltration using a combined pore blockage–cake filtration model. Data were obtained over a range of humic acid concentrations, transmembrane pressures, and stirring speeds. The initial flux decline was due to pore blockage caused by the deposition of large humic acid aggregates on the membrane surface, with a humic acid deposit developing over those regions of the membrane that have first been blocked by an aggregate. The rate of cake growth approaches zero at a finite filtrate flux, similar to the critical flux concept developed for colloidal filtration. The data were in good agreement with model calculations, with the parameter values providing important insights into the mechanisms governing humic acid fouling during microfiltration. In addition, the basic approach provides a framework that can be used to analyze humic acid fouling under different conditions.     

Effect of backwash on the performance of submerged membrane filtration

Particles with a mean diameter of 5 μm were filtered by a ceramic tubular membrane to study the effects of backwash on the performance of submerged membrane filtration. A periodic backwash can completely remove the formed cake, diminishing a part of membrane internal fouling, and, therefore, recover the filtration flux. In a membrane-blocking/cake formation comparable filtration system, the filtration resistance due to membrane-internal fouling is over twice as high as that due to cake formation. The irreversible filtration resistance increases progressively during operation, and it can be regressed to a power-type empirical relationship. Filtration period data were analyzed using blocking models. Membrane blocking occurs in the early filtration periods and is followed by cake filtration. The filtration flux can be simulated by employing blocking models and empirical equations for filtration resistance. The backwash effectiveness was examined by comparing filtrate productivity and washing efficiency. The calculated results of productivity under various backwash durations agree well with experimental data. An increase in backwash flux or duration leads to higher productivity, when the duration is shorter than 2 min; however, the productivity may be decreased with an increase of backwash duration due to the back pumping of more filtrate. A longer filtration time in each cycle results in higher backwash efficiency since a formed cake may efficiently prevent further membrane pore clogging and is more easily removed by a backwash. The optimal backwash conditions can be determined appropriately by the proposed method, with respect to both backwash efficiency and filtrate productivity.     

Modeling of the permeate flux decline during MF and UF cross-flow filtration of soy sauce lees

Cross-flow ultrafiltration and microfiltration have been used to recover refined soy sauce from soy sauce lees for over 25 years. The precise mechanism which dominated the permeate flux during batch cross-flow filtration has not been clarified. In the present study, we proposed a modified analytical method incorporated with the concept of deadend filtration to determine the initial flux of cross-flow filtration and carried out the permeate recycle and batch cross-flow filtration experiments using soy sauce lees. We used UF and MF flat membrane (0.006 m² polysulfone) module under different transmembrane pressures (TMP) and cross-flow velocities. The modified analysis provided an accurate prediction of permeate flux during the filtration of soy sauce lees, because this model can consider the change in J₀ at initial stage of filtration which was caused by the pore constriction and plugging inside membrane, and these changes may not proceed when the cake was formed on the membrane surface. Mean specific resistance of the cake increased with TMP due to the compaction of the cake and decreased with cross-flow velocity due to the change of deposited particle size, but less depended on the membrane in the present study. These results indicate that the value of J₀ determined by modified method was relevant to exclude the effects of the initial membrane fouling by pore constriction due to protein adsorption and plugging with small particles. The modified analytical method for the cake filtration developed in the present study was considered to be capable of selecting an appropriate operating conditions for many cross-flow filtration systems with UF, MF membranes.     

Fabrication of blend hydrophilic polyamide imide (Torlon®)-sulfonated poly (ether ether ketone) hollow fiber membranes for oily wastewater treatment

Blend hydrophilic polyamide imide (PAI)-sulfonated poly (ether ether keton) (SPEEK) hollow fiber membranes were fabricated for oil-water emulsion separation. The structure and performance of the membranes were examined by FESEM analysis, N₂ permeation, overall porosity, collapsing pressure, water contact angle, pure water flux, molecular weight cutoff (MWCO), and oil rejection tests. By studying ternary phase diagrams of polymer/solvent-additive/water system, the higher phase-inversion rate was confirmed for the solutions prepared at higher PAI/SPEEK ratio. A more open structure with larger finger-likes was observed by increasing PAI/SPEEK ratio. Mean pore size of 81 nm, overall porosity of 79% and water contact angle of 58° were obtained for the improved membrane prepared by PAI/SPEEK ratio of 85/15. Increasing SPEEK ratio resulted in lower mechanical stability in terms of collapsing pressure. Pure water flux of about 2.5 times of the plain PAI membrane was found for the improved membrane. MWCO of 460 kDa was found for the improved blend membrane. From oil rejection test, all the membranes demonstrated an oil rejection of over 95%. The improved membrane showed a lower rate of permeate flux reduction compared to the plain membrane which was related to the smaller fouling possibility. Less fouling resistance of the improved membrane was related to the higher flux recovery ratio (about 92%). For all the membranes, the dominant fouling mechanism was found to be the cake filtration. The improved PAI-SPEEK hollow fiber membranes was found to be practical for ultrafiltration of oily wastewaters.     

Membrane performance with a pulp mill effluent: Relative contributions of fouling mechanisms

Severe flux decline was observed during ultrafiltration of a pulp mill effluent. Membrane fouling was the result of varying combinations of adsorption, pore plugging and concentration polarization or gel layer formation. A wide range of membrane materials and pore sizes were evaluated, showing the relationship between the membrane material, pore size and the relative contribution of the different fouling mechanisms. Individual resistances were evaluated for adsorption, R_a, pore plugging, R_pp, and concentration polarization, R_cp, using a series resistance model. These were based on the pure water flux for (1) the new membrane, J_i, (2) after static adsorption with the mill effluent, J_a, (3) the product rate when ultrafiltering the effluent, J_v, and (4) the pure water permeability with the fouled membrane, J_f. These resistances were shown to be misleading in terms of the observed flux loss for cases with significant adsorptive fouling. Adsorptive fouling was underestimated and concentration polarization overestimated. An alternative method, which we shall call flux loss ratios, is proposed, which is based on the flux decline due to a particular mechanisms as a fraction of the overall flux decline. These new measures more accurately reflect the flux decline attributable to each fouling mechanism.     

Modeling of flux decline during crossflow ultrafiltration of colloidal suspensions

Mass transfer during crossflow ultrafiltration is mathematically expressed using the two-dimensional convective–diffusion equation. Numerical simulations showed that mass transfer in crossflow filtration quickly reaches a steady-state for constant boundary conditions. Hence, the unsteady nature of the permeate flux decline must be caused by changes in the hydraulic boundary condition at the membrane surface due to cake formation during filtration. A step-wise pseudo steady-state model was developed to predict the flux decline due to concentration polarization during crossflow ultrafiltration. An iterative algorithm was employed to predict the amount of flux decline for each finite time interval until the true steady-state permeate flux is established. For model verification, crossflow filtration of monodisperse polystyrene latex suspensions ranging from 0.064 to 2.16 μm in diameter was studied under constant transmembrane pressure mode. Besides the crossflow filtration tests, dead-end filtration tests were also carried out to independently determine a model parameter, the specific cake resistance. Another model parameter, the effective diffusion coefficient, is defined as the sum of molecular and shear-induced hydrodynamic diffusion coefficients. The step-wise pseudo steady-state model predictions are in good agreement with experimental results of flux decline during crossflow ultrafiltration of colloidal suspensions. Experimental variations in particle size, feed concentration, and crossflow velocity were also effectively modeled.     

Development of an ultrasonic technique for in situ investigating the properties of deposited protein during crossflow ultrafiltration

Although an amount of research has reported that a flux minimum occurs at the isoionic/isoelectric points (pH 4.6-5.0) in the absence of salts in the ultrafiltration of bovine serum albumin (BSA), the real mechanism remains incompletely understood due to the lack of additional techniques in real time to detect the properties of deposited BSA (gel) layers formed during ultrafiltration (UF). An ultrasonic technique was developed as an analytical noninvasive tool to in situ investigate the properties of deposited BSA layers at pH 4.9 (isoionic or isoelectric point, IEP) and 6.9 during crossflow ultrafiltration. The membrane was a polysulfone (PSf) UF membrane with molecular weight cut-off (MWCO) 35 kDa. The feed used was 0.5 g/l BSA solution. Results show good correspondence between the ultrasonic signal responses and the development of BSA gel layers on the membranes. The deposit is thicker at pH 6.9 than at pH 4.9. However, the deposited gel layers are more compressible at pH 4.9 than at pH 6.9. The flux decline is mainly controlled by the density (packing) of the deposit layer. At pH 6.9, protein mainly deposits on the membrane surface. Around the isoelectric point, protein absorbs within and on the membranes. A functional relationship between acoustic signals and fouling resistance exists. The fouling resistance is mainly attributed to pore blocking or pore constriction.     

Characterization and theoretical analysis of protein fouling of cellulose acetate membrane during constant flux dead-end microfiltration

A rapid characterization method was used to study protein fouling of cellulose acetate membrane during dead-end, in-line, constant flux microfiltration. Based on pressure-permeate volume profiles, two fouling phases could be identified and compared at different permeate fluxes. Using protein staining dyes, the model foulant (bovine serum albumin) was found to deposit on the upstream side of the membrane as a loose cake at its isoelectric point. The effects of solution pH on both the nature and extent of membrane fouling, and membrane cleaning were examined. To further understand and quantitatively analyze the fouling behavior, a combined mathematical model which took into account pore blocking, cake formation and pore constriction was developed based on existing fouling models. The data obtained by modeling was in good agreement with experimental fouling data. Theoretical analysis of data clearly indicated that cake formation was the main fouling mechanism. Using methods such as dynamic light scattering, the significant role of large protein aggregates in membrane fouling was confirmed. The dimer composition of protein did not change significantly during the fouling experiments, clearly indicating that smaller aggregates played less important role in membrane fouling.     

Bacterial attachment on reactive ceramic ultrafiltration membranes

Bacterial attachment is an initial stage in biofilm formation that leads to flux decline in membrane water filtration. This study compares bacterial attachment among three photocatalytic ceramic ultrafiltration membranes for the prevention of biofilm formation. Zirconia ceramic ultrafiltration membranes were dip-coated with anatase and mixed phase titanium dioxide photocatalysts to prevent biofilm growth. The membrane surface was characterized in terms of roughness, hydrophobicity, bacterial cell adhesion, and attached cell viability, all of which are important factors in biofilm formation. The titanium dioxide coatings had minimal impact on the membrane roughness, reduced the hydrophobicity of membranes, prevented Pseudomonas putida attachment, and reduced P. putida viability. Degussa P25 is a particularly promising reactive coating because of its ease of preparation, diminished cell attachment and viability in solutions with low and high organic carbon concentrations, and reduced flux decline. These reactive membranes offer a promising strategy for fouling resistance in water filtration systems.     

Flux decline in crossflow microfiltration and ultrafiltration: experimental verification of fouling dynamics

The theory of fouling dynamics in crossflow membrane filtration is compared with ultrafiltration experiments with suspensions of 0.12 μm silica colloids. It has been experimentally verified that colloidal fouling in crossflow filtration is a dynamics process from non-equilibrium to equilibrium and that the steady state flux is the limiting flux. With the cake concentration c_g identified from an independent experiment and the specific cake resistance calculated by Carman–Kozeny equation, the time-dependent flux and the time to reach steady state in the experiments of this study are correctly predicted with the theory of fouling dynamics.     

Secondary membranes for flux optimization in membrane filtration of biologic suspensions

We employ in situ deposited secondary membranes of yeast (SMYs) to optimize permeate flux during microfiltration and ultrafiltration of protein solutions. The deposited secondary membrane was periodically removed by backflushing, and a new cake layer was deposited at the start of the next cycle. The effects of backflushing time, backflushing strength, wall shear rate, and amount of secondary membrane deposited on the permeate flux were examined. Secondary membranes were found to increase the permeate fluxin microfiltration by severalfold. Protein transmission was also enhanced owing to the presence of the secondary membrane, and the amount of protein recovered was more than twice that obtained during filtration of protein-only solutions under othewise identical conditions. In ultrafiltration, the flux enhancement owing to the secondary membrane was only 50% or less. In addition, the flux for ultrafiltration was relatively insensitive to changes in the concentration of yeast used during deposition of SMY and to the backflushing strength used to periodically remove the secondary membrane.     

Polysilicato-iron for improved NOM removal and membrane performance

The natural organic matter (NOM) removal efficiency of polysilicato-iron (PSI) coagulants and the fouling potential of PSI pretreated waters have been studied using two microfiltration (MF) membrane types: polyvinylidene fluoride (PVDF-2) and polypropylene (PP). The results showed that PSI coagulant with a Si/Fe ratio of 1 (PSI-1) was the most effective, compared to conventional coagulants, in removing dissolved organic carbon (DOC) and in improving the fouling potential. A relative flux of unity through PVDF-2 membrane was achieved for both water sources pretreated with PSI-1.

Aluminium-based coagulants, particularly aluminium chlorohydrate (ACH), worked best at lower coagulant dose. Increasing the coagulant dose to improve DOC removal led to increased membrane fouling, possibly due to increased level of unsettleable flocs and pore blocking. For PSI with larger floc size, the advantage of increased DOC removal was not overridden by the adverse effect of pore blocking. In addition, the residual neutral fraction in the waters and/or the presence of a filter cake on the membrane surfaces seemed to have a limiting effect on the fouling rates through both PP and PVDF-2 membranes to the extent that similar rates were obtained, despite substantial differences in DOC removal.

In contrast, these limiting factors did not influence the fouling potential of PSI-1 treated waters through the PVDF-2 membrane, as suggested by the relative flux of unity for both water sources. It is suggested that the oxide deposits on the PVDF-2 membrane may act as a ‘screening layer’, acting as pre-filtration by the filter cake. This layer may be effectively removed by backwashing, together with deposited NOM, throughout the experiment to maintain the flux at unity. The hydrophobic nature of the PP membrane may discourage the deposition of the oxides, thus minimising the positive effects of the oxides in the system. The high removal of hydrophobic fractions by PSI-1 may also lead to less association between residual NOM and less binding to the membranes, particularly on the PVDF-2 membrane.     

Analysis of particle fouling during microfiltration by use of blocking models

Particle fouling mechanisms in “dead-end” microfiltration is analyzed using blocking models. The blocking index and resistance coefficient of the models during microfiltration are calculated under various conditions. The major factors affecting these model parameters, such as the filtration rate, the amount of particles simultaneously arriving at the membrane surface and particle accumulation, are discussed thoroughly. Instead of the four different blocking models previously proposed, a membrane blocking chart is established for relating the blocking index, filtration rate, and particle accumulation. Blocking index variation during microfiltration can be interpreted using this chart. Membrane blocking occurs during the initial filtration periods until the condition reaches a critical value; then, the blocking index suddenly drops to zero by following up the cake filtration model. Once the normalized resistance coefficient is regressed to an exponential function of the blocking index under a wide range of conditions, the blocking models can be used to quantitatively explain filtration flux attenuation by solving a unitary mathematical equation. Comparing the experimental filtration rates obtained under different conditions with the simulated results reveals a good agreement between them and demonstrates the reliability of this analysis method.