Membrane modules for seawater desalination are becoming increasingly important for obtaining clean water with the rising global water scarcity. The productivity of membrane modules is compromised by biofouling on the membrane and feed spacer. Biofouling development can be mitigated by modification of the spacer or membrane surface. The purpose of the present study is to evaluate the impact of surface-modified feed spacers on the cleaning performance of spiral-wound membrane filtration systems. After cold-plasma treatment, the feed spacers were modified with various combinations of polydopamine (PDA) and silver nanoparticles (AgNP). To compare the cleaning performance of the modified and unmodified spacers, membrane fouling simulators containing nanofiltration membranes and feed spacers were tested under industrially representative conditions: two full cycles involving biofilm development followed by cleaning-in-place (CIP). The modified spacers significantly improved the CIP efficiency when compared with that of the unmodified feed spacer. The highest CIP efficiency was obtained for the PDA–AgNP-coated spacers, which removed >90 % of the biomass. The PDA layers remained undetached during the CIP process, and the amounts of AgNP decreased without affecting the CIP effectiveness during consecutive operational cycles. The results demonstrate that CIP should be included in biofouling tests to evaluate the full potential of surface modifications and suggest that hydrophilic and biocidal spacer surface coatings can significantly improve the CIP effectiveness, thereby considerably reducing the CIP frequency and operational costs.

Early fouling warning is important for the economical operation of membrane separation systems. In parallel multi-channel flow systems, flow re-distribution between channels due to fouling is often associated with maloperation. In the current research we use low magnetic field NMR to monitor multi-fiber hollow fiber membrane modules undergoing a fouling-cleaning cycle and show that rapid detection of fouling is possible by detecting the loss of signal coherence associated with flow re-distribution within the 401 hollow fiber membrane module. This effect is demonstrated to be both reproducible, and reversible via membrane cleaning. The results demonstrate a strong correlation between the coherence signal magnitude and the number of fibers fouled. This may be used in practice for high sensitivity early warning, and to monitor the efficiency of cleaning. This approach may also be particularly useful in the case of detecting residual fouling after cleaning, evidenced in this research by significant flow re-distribution between the before fouling and after cleaning signal coherence.

In this study, a biofouling index based on the relative pressure drop is presented to quantitatively evaluate the amount of fouling in spacer-filled membrane filtration channels. The biofouling index was defined as the inverse of the time to reach a relative pressure drop of 100% and can be interpreted as a fouling rate or cleaning frequency. The index was applied to evaluate biofilm growth in membrane fouling simulators with reverse osmosis membranes and commercial feed spacers operated with different feed water nutrient concentrations and crossflow velocities. Biofilm accumulation on the membrane and feed spacer was characterized in situ using optical coherence tomography. We showed that the biofouling index is directly related to the volume of biofouling independent of the applied crossflow velocity and a suitable tool for improved quantitative comparison of the biofouling rate. Furthermore our results suggest that the pressure drop is better described as function of the velocity at the perimeter of a spacer cell instead of the average velocity in the channel. Although the biofouling index is developed for biofouling, the index may be applied to quantitatively assess mitigation strategies in spacer filled channels for a wider range of fouling types.

In this study non-invasive low field magnetic resonance imaging (MRI) technology was used to monitor fouling induced changes in fiber-by-fiber hydrodynamics inside a multi-fiber hollow fiber membrane module containing 401 fibers. Using structural and velocity images the fouling evolution of these membrane modules were shown to exhibit distinct trends in fiber-by-fiber volumetric flow, with increasing fouling causing a decrease in the number of flow active fibers. This study shows that the fouling rate is not evenly distributed over the parallel fibers, which results in a broadening of the fiber to fiber flowrate distribution. During cleaning, this distribution is initially broadened further, as relatively clean fibers are cleaned more rapidly compared to clogged fibers. By tracking the volumetric flow rate of individual fibers inside the modules during the fouling-cleaning cycle it was possible to observe a fouling memory-like effect with residual fouling occurring preferentially at the outer edge of the fiber bundle during repeated fouling-cleaning cycle. These results demonstrate the ability of MRI velocity imaging to quantitatively monitor these effects which are important when testing the effectiveness of cleaning protocols due to the long term effect that residual fouling and memory-like effect may have on the operation of membrane modules.

The potential of membrane scaling control by a real-time optimization algorithm was investigated. The effect of antiscalant dosing was evaluated from the induction time measured in glass batch-reactors, and from the operational performance of a lab-scale reverse osmosis (RO) unit and two pilot-scale RO units. Step changes in the antiscalant dosing demonstrated that the accumulation of scaling is ‘paused’ during periods when the optimum dose is applied. This is paramount for the application of a dynamic dosing strategy that may briefly underdose, while searching for the optimum dose. It was found that antiscalant underdose and overdose were both detrimental to RO operation since underdose resulted in membrane scaling, while overdose led to membrane fouling due to calcium-antiscalant deposits. The dosing algorithm was used to minimize antiscalant consumption in two pilot RO units. The algorithm was able to lower the antiscalant doses to 0.2 mg/L and 0.6 mg/L, while the supplier's recommended antiscalant doses were 2.0 mg/L and 4.5 mg/L, respectively. As a result, the algorithm could reduce antiscalant consumption by up to 85–90% for the plants mentioned.

The objectives of this study are to assess the performance of antiscalants in increasing the recovery (≥85%) of a reverse osmosis (RO) plant treating anaerobic groundwater (GW) in Kamerik (the Netherlands), and to identify scalants/foulant that may limit RO recovery. Five different commercially available antiscalants were compared on the basis of their manufacturer-recommended dose. Their ability to increase the recovery from 80% to a target of 85% was evaluated in pilot-scale measurements with anaerobic GW and in once-through lab-scale RO tests with synthetic (artificial) feedwater. A membrane autopsy was performed on the tail element(s) with decreased permeability. X-ray photoelectron spectroscopy (XPS) analysis indicated that calcium phosphate was the primary scalant causing permeability decline at 85% recovery and limiting RO recovery. The addition of antiscalant had no positive effect on RO operation and scaling prevention, since at 85% recovery, permeability of the last stage decreased with all five antiscalants, while no decrease in permeability was observed without the addition of antiscalant at 80% recovery. In addition, in lab-scale RO tests executed with synthetic feed water containing identical calcium and phosphate concentrations as the anaerobic GW, calcium phosphate scaling occurred both with and without antiscalant at 85% recovery, while at 80% recovery without antiscalant, calcium phosphate did not precipitate in the RO element. In brief, calcium phosphate appeared to be the main scalant limiting RO recovery, and antiscalants were unable to prevent calcium phosphate scaling or to achieve a recovery of 85% or higher.

The manner in which membrane-fouling experiments are conducted and how fouling performance data are represented have a strong impact on both how the data are interpreted and on the conclusions that may be drawn. We provide a couple of examples to prove that it is possible to obtain misleading conclusions from commonly used representations of fouling data. Although the illustrative example revolves around dead-end ultrafiltration, the underlying principles are applicable to a wider range of membrane processes. When choosing the experimental conditions and how to represent fouling data, there are three main factors that should be considered: (I) the foulant mass is principally related to the filtered volume; (II) the filtration flux can exacerbate fouling effects (e.g., concentration polarization and cake compression); and (III) the practice of normalization, as in dividing by an initial value, disregards the difference in driving force and divides the fouling effect by different numbers. Thus, a bias may occur that favors the experimental condition with the lower filtration flux and the less-permeable membrane. It is recommended to: (I) avoid relative fouling performance indicators, such as relative flux decline (J/J0); (II) use resistance vs. specific volume; and (III) use flux-controlled experiments for fouling performance evaluation.

The role of phosphate and humic substances (HS) in preventing calcium carbonate scaling and their impact on antiscalant dose was investigated for a reverse osmosis (RO) system treating anaerobic groundwater (GW) (containing 2.1 mg/L orthophosphate and 6-8 mg/L HS). Experiments were conducted with the RO unit (treating anaerobic GW), and with a once-through lab-scale RO system (operating with artificial feedwater). Additionally, (batch) induction time (IT) measurements were performed with, i) real RO concentrate, and ii) artificial RO concentrates in the presence and absence of phosphate and HS. It was found that at 80% recovery (Langelier saturation index (LSI) 1.7), calcium carbonate scaling did not occur in the RO unit when the antiscalant dose was lowered from 2.2 mg/L (supplier's recommended dose) to 0 mg/L. The IT of the real RO concentrate, without antiscalant, was longer than 168 h, while, at the same supersaturation level, the IT of the artificial concentrate was approximately 1 h. The IT of the artificial concentrate increased to 168 h with the addition of 10 mg/L of phosphate, humic acid (HA), and fulvic acid (FA). Furthermore, in the lab-scale RO tests, the normalized permeability (Kw) of the membrane decreased by 20% in 2 h period when fed with artificial concentrate of 80% recovery containing no phosphate, whereas, with phosphate, no decrease in Kw was observed in 10 h period. These results indicate that phosphate and HS present in the GW prevented calcium carbonate scaling in the RO unit and reduced the use of commercial (synthetic) antiscalants.

In this study, we analyzed the heat requirement of membrane distillation (MD) to investigate the trade-off between the evaporation efficiency and driving force efficiency in a single effect MD system. We found that there exists a non-zero net driving temperature difference that maximizes efficiency. This is the minimum net driving temperature difference necessary for a rational operational strategy because below the minimum net driving temperature, both the productivity and efficiency can be increased by increasing the temperature difference. The minimum net driving temperature has a similar magnitude to the boiling point elevation (~0.5 °C for seawater), and depends on the properties of the membrane and the heat exchanger. The minimum net driving temperature difference concept can be used to understand the occurrence of optimal values of other parameters, such as flux, membrane thickness, and membrane length, if these parameters are varied in a way that consequently varies the net driving temperature difference.

In this study at a full-scale desalination installation seven types of commercially available cartridge filter (CF) elements were evaluated in terms of: (i) water production volume (m³), (ii) produced water quality, and (iii) operational cost (€cent/m³). The cost of optimal CF replacement time relative to increased CF pressure drop was determined for three electricity tariffs (0.05, 0.15, and 0.25 €/kWh) to assess further cost reduction. CF 1 was able to achieve the highest water production rate, the lowest produced water SDI, and the lowest cost of operation. The total costs of cartridge filtration varied between 1.22 and 1.70 €cents/m³ produced water, depending on the CF type. Replacing the worst-performing CF type by the best-performing CF type would reduce operational CF costs by about 39.3%, enabling a cost saving of 0.48 €cents/m³ produced water, emphasizing that selection of the right CF enables a large reduction of cartridge filtration costs. Moreover, depending on the electricity tariff an additional 2–16% cost reduction can be achieved by replacing CFs at an optimal time. At high energy cost, it may be more economical to replace cartridge elements more often to reduce the increased cost associated with the required higher pressure.

High pressure membranes are increasingly used for the treatment of contaminated water for various purposes including irrigation and drinking water. The lack of a fast and easy to implement membrane integrity test method with a log removal value (LRV) >3 hampers the implementation of these membranes. Current on-line methods include conductivity, TOC (total organic carbon) and turbidity measurements and can monitor a maximum LRV of 3. Furthermore, challenge tests using chemical or bacteriological virus surrogates such as bacteriophage MS2 show that RO and NF systems can reach LRVs of 6-7, but dosing of these surrogates is not feasible and desirable in full scale drinking water plants. This study describes the identification and use of indigenous viruses, naturally present in surface waters to monitor the integrity of RO membranes in a pilot installation. Natural viruses were identified from fresh source water using metagenomics and qPCR primers developed for a selected set of viruses that were present in high numbers in surface water. The qPCR assays were used to determine the number of gene copies of these viruses in the feed and permeate of the pilot RO installation, and the LRV of these natural viruses was compared with the LRV of spiked MS2 and with on-line conductivity. The concentration of the selected natural viruses in the source water was sufficient to demonstrate a LRV of >7 and was comparable to the results of the spiked MS2 bacteriophage. Furthermore, after inflicting damage to the membrane element by drilling small holes of 1 and 4 mm, both MS2 and the natural viruses detected the damage to the membrane with a nearly identical decrease of LRV, while conductivity lacked sensitivity to monitor any integrity loss. This novel method enables monitoring of the RO membrane integrity at a high sensitivity (LRV > 7), without the addition of chemical or biological virus surrogates. Furthermore, the high concentration of viruses in source water simplifies detection without laborious sample concentration procedures. The implementation of this method facilitates monitoring of the integrity of RO membranes in full scale operation with a much higher sensitivity than current methods.

Ensuring the biological stability of drinking water is essential for modern drinking water supply. To understand and manage the biological stability, it is critical that the bacterial growth in drinking water can be measured. Nowadays, advance treatment technologies, such as reverse osmosis (RO), are increasingly applied in drinking water purification where the produced water is characterized by low levels of nutrients and cell counts. The challenge is, therefore, how to measure the low bacterial growth potential (BGP) of such ultra-pure water using the available methods which were originally developed for conventionally treated drinking water. In this study, we proposed a protocol to assess BGP of ultra-pure drinking water produced by RO and post-treatment (including remineralization). Natural bacterial consortium from conventional drinking water was added to all water samples during this study to ensure the presence of a wide range of bacterial strains. The method development included developing an ultra-pure blank with high reproducibility to lower the detection limit of the BGP method (50 ± 20 × 10³ intact cells/mL) compared with conventional blanks such as bottled spring water, deep groundwater treated by aeration and slow sand filtrate of surface water supply. The ultra-low blank consists of RO permeate after adjusting its pH and essential mineral content under controlled laboratory conditions to ensure carbon limitation. Regarding the test protocol, inoculum concentrations of >10 × 10³ intact cells/mL may have a significant contribution to the measured low levels of BGP. Pasteurization of water samples before measuring BGP is necessary to ensure reliable bacterial growth curves. The optimized method was used to assess BGP of ultra-pure drinking water produced by RO membranes and post-treatment (including remineralization), where the BGP has decreased more than 6-fold to a level of 90 ± 20 × 10³ intact cells/mL compared with conventionally treated water (630 ± 70 × 10³ intact cells/mL).

Depletion of fresh groundwater sources as the result of overdraft, salinization and pollution becomes a major problem in parts of the world. Desalination of brackish groundwater by membrane technology, e.g. reverse osmosis (RO), seems to be a promising solution to water scarcity problems. However, energy consumption and concentrate disposal are considered as the main reasons for avoiding RO application. In order to overcome these drawbacks, the PURO concept, which consists of vertically-configured RO unit in an especially drilled well is designed, installed and is going to be tested. The installation operates without any chemical pretreatment and therefore, the concentrate can be injected into a deeper confined aquifer that contains water of similar concentration. To avoid chemical pretreatment, the system operates at lower recovery (50%) than conventional brackish groundwater reverse osmosis (BWRO). Higher energy consumption, as the results of lowering the recovery, is avoided by using natural hydrostatic pressure at the depth that RO is installed and by extracting the permeate water only. PURO consumes about 39% less energy when compared to a conventional BWRO installation of the same capacity. This article describes the PURO concept and discusses its advantages and disadvantages. It also provides a rough calculation of water cost for PURO and conventional BWRO with emphasizing on the energy costs.