A concise analytical model for the ideal reverse osmosis desalination processes.

Permeate recovery is a key parameter that plays a central role in the performance assessment and optimization of reverse osmosis processes. It remains a great challenge for engineers in the field to determine the recovery conveniently and accurately from the basic parameters of a membrane system. A concise analytical model is presented here that, without the need for the empirical or fitting coefficients and tedious numerical calculation, links the recovery of a reverse osmosis process rigorously to three basic parameters: the feed water salt concentration, the characteristic of membrane process, and the driving pressure. A graphical solution method to the model is also introduced to find out the recovery effortlessly. The concise model and graphical method are demonstrated under various conditions as a powerful tool for the performance simulations and improvement of reverse osmosis processes. PRACTITIONER POINTS: A concise model is presented for permeate recovery in an ideal RO desalination system. The recovery of an RO system is analytically related to a few system parameters. A graphical solution method is introduced to find out the explicit recovery. The ideal RO model can be a powerful tool for simulation of RO processes.

Water transport in reverse osmosis membranes is governed by pore flow, not a solution-diffusion mechanism.

We performed nonequilibrium molecular dynamics (NEMD) simulations and solvent permeation experiments to unravel the mechanism of water transport in reverse osmosis (RO) membranes. The NEMD simulations reveal that water transport is driven by a pressure gradient within the membranes, not by a water concentration gradient, in marked contrast to the classic solution-diffusion model. We further show that water molecules travel as clusters through a network of pores that are transiently connected. Permeation experiments with water and organic solvents using polyamide and cellulose triacetate RO membranes showed that solvent permeance depends on the membrane pore size, kinetic diameter of solvent molecules, and solvent viscosity. This observation is not consistent with the solution-diffusion model, where permeance depends on the solvent solubility. Motivated by these observations, we demonstrate that the solution-friction model, in which transport is driven by a pressure gradient, can describe water and solvent transport in RO membranes.

Characterization of activated sludge flocs in membrane bioreactor: stable and unstable flocs.

In this study, the properties of unstable and stable flocs were investigated under the steady operation of a membrane bioreactor (MBR). The extracellular polymeric substances (EPS) composition, surface charge, and hydrophobicity of unstable and stable flocs were examined and compared. Interfacial interactions of the membrane with unstable flocs, unstable flocs themselves, and unstable and stable flocs were assessed using the extended Derjaguin-Landau-Verwey-Overbeek (XDLVO) models. Cake layer resistance was found to contribute more than 80% of total resistance under steady operating conditions. Compared with stable flocs, unstable flocs possessed a higher level of EPS, more diverse protein, more negative charge, weaker hydrophobicity, and higher fouling potential. Thermodynamic analyses showed that unstable flocs had a higher adhesive strength (- 63.4 mJ/m2) with the membrane, lower self-cohesive strength (- 18.3 mJ/m2), and higher cohesive strength (- 54.3 mJ/m2) with stable flocs. Therefore, some unstable flocs remained on the membrane surface to form the cake layer due to their poor cohesion strength.

Characterization of the Fouling Layer on the Membrane Surface in a Membrane Bioreactor: Evolution of the Foulants' Composition and Aggregation Ability.

In this study, the characteristics of membrane foulants were analyzed with regard to morphology, composition, and aggregation ability during the three stages of transmembrane pressure (TMP) development (fast-slow-fast rise in TMP) in a steady operational membrane bioreactor (MBR). The results obtained show that the fouling layer at the slow TMP-increase stage possessed a higher average roughness (71.27 nm) and increased fractal dimension (2.33), which resulted in a low membrane fouling rate (0.87 kPa/d). A higher extracellular DNA (eDNA) proportion (26.12%) in the extracellular polymeric substances (EPS) resulted in both higher zeta potential (-23.3 mV) and higher hydrophobicity (82.3%) for initial foulants, which induced and increased the protein proportion in the subsequent fouling layer (74.11%). Furthermore, the main composition of the EPS shifted from protein toward polysaccharide dominance in the final fouling layer. The aggregation test confirmed that eDNA was essential for foulant aggregation in the initial fouling layer, whereas ion interaction significantly affected foulant aggregation in the final fouling layer.

Metastable State of Water and Performance of Osmotically Driven Membrane Processes.

Semipermeable membranes play critical roles in many natural and engineering systems. The osmotic pressure is found experimentally much less effective than the hydraulic pressure to drive water through the membrane, which is commonly attributed to the internal concentration polarization (ICP) in the porous layer of the membrane. In this study, it has been shown that a necessary condition for the osmotic pressure to be effective is water continuity across the entire membrane thickness under negative pressure, i.e., the water inside the membrane remains in a metastable state. However, the metastable state of water cannot be maintained indefinitely, and cavitation will undoubtedly occur in the osmotically driven processes. Collapse of the water metastable state was suggested for the first time to be a more important and fundamental reason for the low water fluxes in the osmotically driven membrane processes.

Dynamic analysis of self-forming dynamic membrane (SFDM) filtration in submerged anaerobic bioreactor: Performance, characteristic, and mechanism.

This study attempts to provide an improved fundamental understanding of the self-forming dynamic membrane (SFDM) filtration process in submerged anaerobic bioreactors. Excellent system performances were achieved in terms of high COD removal efficiency (∼ 90%), fast formation/reformation of SFDM (<1 h), and sustainable low-resistance (3.92 × 1010 m-1) high-flux (10-30 L/m2·h) filtration. A typical flux-variation profile consisted of an initial abruptly fast decrease followed by a gradually slow reduction, corresponding to the formation and sustainable operation period, respectively. The increase of SFDM resistance in formation period was attributable to the fast deposition of large particles on coarse-pore support materials. After SFDM formation, the subsequent increase of SFDM resistance was controlled more by the increase of specific resistance, which was firstly mainly resulted from the increasing accumulation of small particles with higher hydrophobicity and the external deposition of eEPS but later most attributable to the increase of internal release of eEPS.

Micro-bubbles enhanced breakage warning for hollow fiber membrane integrity with a low-cost real-time monitoring device.

A low-cost device was developed to monitor the integrity of hollow fiber membrane by real-time online detecting and measuring air bubbles in the permeate flow. When a breakage occurs in the fiber membrane system, air bubbles will find their ways to enter the permeate flow. The monitoring device consists of two pairs of platinum probes attached to a pipe, along which the permeate flows. Membrane's breakage was identified by the voltage changes between the two pairs of probes when air bubbles in the permeate touch them. There was no addition of any other chemicals or materials into the system that would jeopardize final products of the membrane process. Experimental results showed that the voltage signal changes before and after a breach of membrane were obvious in the normal operation conditions. The smallest diameter of the bubbles that can be detected by the monitoring device was 50 µm, which was captured and identified with a high-speed camera. Furthermore, the sensitivity of device with two pairs of probes improved by 5~9% over that with one pair of probes. Finally, cost of the proposed device was roughly estimated only about 70 US dollars and the detection result was highly consistent with that obtained with the prevailing particle counters of several thousand US dollars.

Decolorization and Mineralization of Rhodamine B in Aqueous Solution with a Triple System of Cerium(IV)/H2O2/Hydroxylamine.

Hydroxylamine (HA) can react with hydrogen peroxide (H2O2) to generate hydroxyl radical (HO•), but the reaction rate between them is very slow (2.2 × 10-4 M-1 s-1). We propose a new system to accelerate the formation of aminoxyl radical (NH2O•) by the addition of cerium [Ce(IV)] to induce the continuous production of HO• through reaction with H2O2. We also investigate the decolorization and mineralization of rhodamine B (RhB) and mechanism in the Ce(IV)/H2O2/HA system. The initial pH plays a significant role in decolorization of RhB. In this work, observation of the rapid decolorization process after 60 min revealed that approximately 80% of RhB was degraded at the initial pH of 4.0. The HO• radicals were considered as the primary reactive oxidant in the system, during its investigation through coumarin capturing, benzoic acid capturing, and radical quenching experiments. The results of the present study suggest that the addition of Ce(IV) can greatly enhance the production of HO•, and the rapid decolorization and mineralization of RhB can occur through the Ce(IV)/H2O2/HA system at acidic pH conditions.

Bisection method for accurate modeling and simulation of fouling in hollow fiber membrane system.

Accurate description and modeling of fouling on hollow fibers imposes a serious challenge to more effective fouling mitigation and performance optimization of the membrane system. Although the governing equations for membrane fouling can be constructed based on the known theories from membrane filtration and fluid dynamics, they are unsolvable analytically due to the complex spatially and temporally varying nature of fouling on hollow fibers. The current available numerical solutions for the governing equations are either unreliable or inconvenient to use because of the uses of unfounded assumptions or cumbersome calculation methods. This work presented for the first time a rigorous numerical procedure to solve the governing equations for fouling development on hollow fibers. A critical step to achieve the goal is the use of bisection method to determine the transmembrane pressure at the dead end of the fibers. With this procedure, fouling behavior in the hollow fiber membrane system under a given condition can be simulated within a second. The model simulations were well calibrated and verified with the published experimental data from literature. Also presented in the paper were simulations for performances of the hollow fiber membrane system under various operation conditions. Graphical abstract ᅟ.

Calcium ion on membrane fouling reduction and bioflocculation promotion in membrane bioreactor at high salt shock.

Fouling propensity of activated sludge in membrane bioreactor (MBR) is closely related to the disturbance of a salt shock. In this work, the characteristics of membrane fouling and bioflocculation were compared in two laboratory-scale MBRs (one with calcium addition, MBR-Ca, the other without, MBR-C) with a transient salt shock. Particle size distributions, zeta potential, relative hydrophobicity, modified fouling index, the content of polysaccharides, proteins and calcium ions in different layers of sludge were monitored prior to, during and after the salt shock. Comparison with MBR-C showed that the recovery time and fouling rate of MBR-Ca were reduced by 50% and 34%, respectively. Remarkable variations of sludge properties in terms of bioflocculation, such as larger particle sizes, higher relative hydrophobicity and zeta potential, lower polysaccharides in supernatant, higher proteins/polysaccharides ratio in slime and loose bound extracellular polymeric substances, were observed in MBR-Ca after the salt shock.

Stratification structure of polysaccharides and proteins in activated sludge with different aeration in membrane bioreactor.

The effect of distribution pattern of polysaccharides (PS) and proteins (PN) in activated sludge (AS) stratification with different aeration rates on membrane fouling and rejection efficiency were investigated. During high aeration, PN and PS concentrations increased in supernatant, the dominant fraction (84% of PN and 73% of PS) was small molecules (<1 kDa). Less slime and loose bound extracellular polymeric substances (LB-EPS), more tight bound EPS (TB-EPS) were observed compared with low aeration. The decrease in PN/PS ratio and Ca(2+) concentration within EPS deteriorated AS flocculation ability. At slow trans-membrane pressure (TMP) rise stage, fouling rate under high aeration was 41% lower than low aeration due to lower PN within EPS outer. Low PS rejection rate (about 23%) leaded to higher PS in effluent at this stage. High PS rejection rate (about 94%) at rapid TMP rise stage resulted in about 2.2-time higher fouling rate than that low aeration.

Impact of sludge cation distribution pattern on its filterability in membrane bioreactor.

The distributions of cations of various valences (Na(+), Ca(2+) and Fe(3+)) in the outer layers of extracellular polymeric substances (EPSs) to pellet have a significant impact on the stratification structure of polysaccharides (PS) or proteins (PN) in activated sludge. Comparison with the control showed that the monovalent Na(+) reduced flocculability slightly (about 9.75%), whereas Ca(2+) and Fe(3+) increased flocculability significantly. The modified fouling index (MFI) had a significant correlation with PN in the supernatant (rp=0.8593), slime (rp=0.7218) and loosely bound EPS (LB, rp=0.8012). However, it had a moderate correlation with PS in supernatant (rp=0.5842), and weak correlation to slime (rp=0.3785) or LB (rp=0.3219). There was an ignored correlation with PN or PS in the tightly bound EPS (TB) or pellet. The lower amount of PN or PS in the supernatant would have positive impact on improving the activated sludge filterability.

Performance enhancement and fouling mitigation by organic flocculant addition in membrane bioreactor at high salt shock.

The main objective of this study was to investigate the effect of an organic flocculant (MPE50) addition on reducing membrane fouling and enhancing performance in membrane bioreactor (MBR) at the high salt shock. Results show that MPE50 addition is a reliable and effective approach in terms of both membrane fouling mitigation and pollutants removal improvement in the case of high salt shock. Compared to the control reactor, the MBR with MPE50 addition enhanced the average removal of COD, NH4(+)-N and TP by 4.1%, 13.2% and 21.2%, respectively. Due to the effect of flocculation and adsorption by MPE50, a significant reduction in the soluble microbial products (SMP) proteins amount was observed. As a result, the membrane fouling rate was mitigated successfully. Further, the increasing of mean particles size, Zeta potential and related hydrophobicity of the flocs would also have positive impacts on membrane fouling mitigation.

Modeling of concentration polarization in a reverse osmosis channel with parabolic crossflow.

Concentration polarization in narrow reverse osmosis channels with parabolic crossflow was numerically simulated with finite different equations related to permeate velocity, crossflow velocity, average salt concentration, and wall salt concentration. A significant new theoretical development was the determination of two correction functions, F2 and F3, in the governing equation for average salt concentration. Simulations of concentration polarization under various conditions were then presented to describe the features of the new model as well as discussions about the differences of concentration polarizations of the more realistic parabolic flow with those when plug flow or shear flow was assumed. The situations in which the simpler models based on shear or plug flow can be used were indicated. Concentration polarization was also simulated for various conditions to show the applicability of the model and general features of concentration polarization in a narrow, long reverse osmosis channel.

Two-step optimization of pressure and recovery of reverse osmosis desalination process.

Driving pressure and recovery are two primary design variables of a reverse osmosis process that largely determine the total cost of seawater and brackish water desalination. A two-step optimization procedure was developed in this paper to determine the values of driving pressure and recovery that minimize the total cost of RO desalination. It was demonstrated that the optimal net driving pressure is solely determined by the electricity price and the membrane price index, which is a lumped parameter to collectively reflect membrane price, resistance, and service time. On the other hand, the optimal recovery is determined by the electricity price, initial osmotic pressure, and costs for pretreatment of raw water and handling of retentate. Concise equations were derived for the optimal net driving pressure and recovery. The dependences of the optimal net driving pressure and recovery on the electricity price, membrane price, and costs for raw water pretreatment and retentate handling were discussed.

Impact of feed water acidification with weak and strong acids on colloidal silica fouling in ultrafiltration membrane processes.

Although acidification of feed water is a common practice to prevent scaling of the sparingly soluble minerals in nanofiltration and reverse osmosis processes, the change of acidity may have a potentially adverse impact on colloidal fouling, which is another important type of fouling on the membranes. In this paper, commonly used strong and weak acids are quantitatively investigated for their effect on colloidal silica fouling with a lab-scale ultrafiltration (UF) membrane system. Experiments showed that addition of either strong or weak acids in feed water would intensify colloidal fouling. However, the strength of colloidal fouling with strong acid addition was consistently higher (12-37%) than that with weak acid addition at pH 3. The smaller increase in colloidal fouling potential observed with weak acids was attributed to the adsorption of weak acid anions on the colloidal silica surface, which kept the absolute value of zeta potential of the colloids relatively high. Consequently, the difference in colloidal fouling potential with the additions of strong and weak acids diminished at high salt concentration. The findings implied that the type of acid used in feed water acidification could have a significant impact on colloidal fouling for low-salinity waters.

Differential pressure in membrane channel caused by foulant capture onto spacers.

Feed spacers in spiral-wound reverse osmosis membrane modules are highly susceptible to deposition or attachment of suspended particulates and organic matters in the feed stream. This type of fouling can cause a significant pressure drop along the membrane channel (differential pressure) without much effect on the average permeate flux. In practical applications, membrane cleaning is often triggered when the differential pressure in a membrane channel exceeds a threshold value. A mathematical model was developed to simulate the development of differential pressure in the feed channel resulting from foulant capture by the feed spacers. Simulations were carried out to investigate and demonstrate the effect of various parameters on differential pressures in the membrane channel. Differential pressures observed in a two-stage reverse osmosis water reclamation plant were simulated with the model, and the results showed that the model could adequately describe the increase of differential pressure with operating time in the reverse osmosis plant.

Soluble microbial products in membrane bioreactor operation: Behaviors, characteristics, and fouling potential.

This paper presents an experimental study on soluble microbial products (SMP) in membrane bioreactor (MBR) operation at different sludge retention times (SRTs). A laboratory-scale MBR was operated at SRT of 10, 20, and 40 days for treatment of readily biodegradable synthetic wastewater. The accumulation, composition, characteristics, and fouling potential of SMP at each SRT were examined. It was found that accumulation of SMP in the MBR became more pronounced at short SRTs. Carbohydrates and proteins appeared to be the components of SMP prone to accumulate in the MBR compared with aromatic compounds. The proportions of SMP with large molecular weight in supernatants and in effluents were almost identical, implying that membrane sieving did not work for most SMP. In addition, the majority of SMP was found to be composed of hydrophobic components, whose proportion in total SMP gradually increased as SRT lengthened. However, fouling potentials of SMP were relatively low at long SRTs. The hydrophilic neutrals (e.g., carbohydrates) were most likely the main foulants responsible for high fouling potentials of SMP observed at short SRTs.

A modeling study of fouling development in membrane bioreactors for wastewater treatment.

Membrane fouling is a primary concern in membrane bioreactors (MBRs) in wastewater treatment because it strongly affects both system stability and economic feasibility. A mathematical model was developed in this study for membrane fouling in submerged MBR systems for wastewater treatment, in which both reversible and irreversible fouling were quantified. While mixed liquor suspended solids are the major components of the reversible fouling layer, dissolved organic matter is thought to be the key foulant, in particular, responsible for the long-term irreversible fouling of the filtration unit. The model was calibrated (parameter identification) with a set of operational data from a pilot MBR and then verified with other independent operational data from the MBR. The good agreement between theoretical predictions and operational data demonstrates that the outlined modeling concept can be successfully applied to describe membrane fouling in submerged MBR systems.

Influence of various monovalent cations and calcium ion on the colloidal fouling potential.

The influence of various monovalent cations and of divalent calcium ions on colloidal fouling strength was investigated quantitatively on a bench-scale ultrafiltration device. A higher colloidal fouling potential (k) was consistently observed with lithium chloride compared to the same ionic strengths of chlorides of other monovalent cations (Na+, K+, and Cs+). This observation was attributed to the formation of an impervious layer around the colloidal particle by lithium ions that prevented the repulsive forces due to the interaction of the silica hairs formed on the particles in the presence of water. The impact of the divalent calcium ion on the fouling potential was more complex. The fouling potential first increased with calcium ion concentration and then decreased. The maximum value of fouling potential occurred at the ionic strength corresponding to the critical coagulation concentration, which decreased with increasing colloid concentration. The colloidal fouling potential was well correlated by a bilinear relationship with colloid concentration and ionic strength for all salts tested under the critical coagulation concentration.