Classic conventional pretreatment processes might consist of all the following treatment strategies or just some of them, including pre-screening, chlorination, coagulation and flocculation, sedimentation/dissolved air flotation (DAF), granular media filtration, ozonation, and scale inhibitors.

The pre-screening process is the first and most basic important step in the RO water treatment process. The purpose of this pretreatment is to remove large and non-soluble solids to reduce the pollutant load and protect all the subsequent treatment steps. Large particles in the raw water of the treatment plant, including plants, fish, seashells, and microorganisms, could be attached to and grow in the water intake pipelines. Therefore, prior to further processing, the raw water should be screened to remove these materials. A simple mesh inclined at a specific angle serves as a screening process to block the intrusion of marine creatures at the intake grids. Then, a mechanical rake attached to the screens removes them from the mesh ^[1][63]. Although different types of screens have been suggested in the literature, the appropriate screen should be selected based on the characteristics of the feed water, site requirements, and hydraulic calculations ^[2][86]. Factors such as the size of the screening openings and mechanical properties are commonly used to classify the screens. The screening process in desalination treatment plants often employs screens with openings between 120 and 500 µm. Previous studies have shown that screens with openings between 400 to 500 µm perform well in treating fresh surface source water and provide an effective screening solution prior to membrane filtration ^[3][57]. However other studies have reported that screens with openings larger than 120 µm may not be effective for the seawater source. Seawater contains barnacle larvae, which can pass through larger openings and damage the pretreatment process as they can withstand chlorination. Therefore, Murkute et al. have revealed that the use of a finer screen filter (100 µm) is necessary for removing barnacle larvae if MF or UF pretreatment is selected for open seawater intakes ^[4][87]. It is worth mentioning that granular media filtration pretreatment systems do not require micro-screens since they can remove barnacles at all stages of development. Therefore, they are adequately protected by traveling screens of 3 to 10 mm. When comparing conventional granular media filtration and membrane filtration pretreatment, it is important to consider the cost of micro-screening systems ^[3][57].

Conventional water disinfection is often achieved through chlorination to inhibit biological growth that leads to filter and membrane biofouling. Pre-chlorination consists of adding chlorine to the raw water after it has been screened ^[5][49]. Chlorine is a commonly used disinfectant in water desalination because it is readily available, easy to apply, and can effectively deactivate many microorganisms. As an oxidant, it destroys cellular, microbiological, and internal components of microorganisms ^[6][88]. Three specific mechanisms have been proposed for chlorine-based bacterial inactivation ^[7][89]. For example, Fiedler et al. found that chlorinated effluents significantly reduced membrane fouling due to the partial oxidation and the change in the organic matter’s properties, which affects its reactivity with membranes ^[8][90]. There are several chemicals that can be used in pre-chlorination. Chlorine gas and sodium hypochlorite are perhaps the most commonly used forms of chlorine that form hypochlorous acids in water. Gaseous chlorine is more cost-effective than hypochlorite. However, since it is more dangerous than hypochlorite, it needs careful handling ^[9][91]. Shock chlorination and continuous chlorination could be used in water pretreatment. It has been shown that shock chlorination was better than continuous chlorination since long-term exposure to chlorine could destabilize natural organic matter such as natural colloidal polymers, which further promotes more coagulation of these compounds ^[10][92]. In addition, continuous chlorination increased the concentration of some foulants such as polysaccharides and irritated sea organisms ^[5][49]. However, it is worth noting that the application mode (continuous or shock), as well as the optimal chlorine dosage, are mainly site-specific, depending on different factors such as environmental conditions, local regulations, permeate quality, and the existence of other interfering substances such as transition metals ^[7][89]. Although chlorination is the most effective method for disinfection and odor control, chlorination has some serious drawbacks. Carcinogenic by-products such as trihalomethanes and haloacetic acids could be formed from its reaction with organics in water ^[11][60]. In addition, polyamide membranes’ sensitivity to chlorine oxidation has led the membrane to be susceptible to biofouling after dechlorination using sodium metabisulfite, as the residual chlorine has to be eliminated prior to the RO membranes to prevent their damage ^[7][89]. These have intensified the search for chemical alternatives. Chloramine, peroxide, and chlorine dioxide have been proposed as oxidizing chemicals, while isothiazolones and 2,2-di-bromo-3-nitrioproprionamide have been suggested as non-oxidizing chemical alternatives ^[12][93]. Nevertheless, these chemicals are not as effective as chlorine and lack many characteristics of the ideal disinfectant given by Bates ^[7][89]. Membrane modification and the development of anti-fouling membranes are promising techniques to mitigate biofouling post-dichlorination and/or increase chlorine resistance ^[13][14][15][94,95,96]. Consistent with this idea, a lab. scale study conducted by Hong et al. ^[15][96] showed promising chlorine-resistant polyamide membrane production to overcome the severe issue of biofouling. However, there are challenges that require further research to scale up these alternative techniques ^[16][69].

Coagulation–flocculation processes are utilized to enhance the elimination of suspended solids and colloidal particles from water ^[17][97]. They can also be used to remediate some bacteria and dissolved organic matter ^[18][98]. It is used in the first stage of solids–liquids separation. The process typically encompasses the addition of coagulant chemicals to destabilize colloidal particles and form larger, heavier particles that can be separated easily using sedimentation, flotation, or filtration. Common coagulants used include ferric salts, such as ferric sulfate and ferric chloride, and aluminum salts, including aluminum and polyaluminum chloride ^[1][63]. The usual dose for inorganic coagulants is between 5 and 30 mg/L, while polymers often require only 0.2–1 mg/L ^[19][99]. The pH of the water plays a crucial role in coagulation. Iron salts are typically preferred to aluminum salts due to difficulties in controlling the pH and potential scaling problems in RO membranes ^[5][20][49,100]. Ferric-based coagulants, particularly ferric chlorides, are commonly used in desalination plants. For example, Fujairah II, the largest hybrid desalination plant in the world, uses ferric chloride in coagulant tanks before filtration ^[21][101]. This same coagulant is also used in a desalination plant in Saudi Arabia to enhance the performance of the subsequent dual-media filtration and to mitigate biofouling. To improve the coagulation stage and prevent calcium carbonate scaling formation, an acid solution is added alongside the coagulant to lower the pH of the feed water ^[22][102]. Coagulation using ferric chloride has been demonstrated to be effective in eliminating suspended solids, colloidal particles, and natural organic matter such as humic and fulvic acids and algal organic matter ^[23][24][103,104]. Studies have also found that using a ferric-based coagulant can reduce the concentration of algal organic matter in seawater when combined with UF. Low dosages of coagulant are sufficient to decrease the fouling potential of the membranes, but iron-biopolymer aggregates, produced by the adsorption of biopolymers to iron hydroxide, can reduce the flux-dependency of algal organic matter fouling ^[23][103]. Additionally, it has been shown that residual iron can have negative effects on both RO membranes ^[20][100] and thermoelectric plants by causing corrosion ^[25][105]. To prevent this, polyaluminum coagulants have been tested as an alternative to conventional inorganic coagulants. One such coagulant, polyaluminum chloride, has been shown to not consume excessive amounts of alkalinity, resulting in little variation in pH when added to water ^[26][27][106,107]. However, it is crucial to note that these synthetic polymers are toxic, and their monomers are carcinogenic ^[19][99]. Therefore, researchers are testing different alternative coagulants to mineral coagulants. Cationic organic compounds may be able to replace inorganic coagulants in some cases as they can directly neutralize negative colloids ^[11][60]. Alshahri et al. ^[28][108] have proposed the use of clays combined with low doses of liquid ferrate as an alternative coagulant that can effectively remove turbidity, dissolved organic carbon, algal organic matter, and also reduce chemical consumption and sludge production. However, more research is needed to investigate the application of this strategy on a large scale. Duan et al. ^[29][109] found that the use of powdered activated carbon, before the addition of the metal salt coagulant, is recommended to significantly reduce the concentration of humic acid. It is worth noting that coagulation has been proven to be an effective technique for enhancing water quality in both conventional and membrane filtration pretreatment technologies, which could surpass conventional pretreatment in terms of performance ^[30][110]. Therefore, the focus of current and future research should be on integrating conventional technologies, particularly coagulation–flocculation, with low-pressure membrane pretreatment technologies.

Sedimentation and DAF are common water pretreatment technologies utilized to separate liquid and solid phases based on density and buoyancy, respectively, after coagulation–flocculation and upstream of a granular media filter. These two clarification methods are used to produce clarified water upstream of the granular media filter.

Sedimentation is the process of allowing the flocculate particles, formed after coagulation and flocculation, to settle at the bottom of a sedimentation tank under the influence of gravity following an optimal detention period. The settled particles are then pumped out of the system through a sludge pipeline ^[31][111]. The main purpose of sedimentation is to reduce the total suspended solids (TSS) concentration to improve the efficiency of subsequent filtration while avoiding the need for continuous backwashing. The efficiency of the sedimentation system is influenced not only by TSS concentration but also by the volume/area of the tank, the flow rate through the tank, and the settling velocity of the suspended particles ^[32][33][112,113]. Sedimentation can effectively remove suspended solids when source water has a daily average turbidity of higher than 30 NTU (TSS concentration higher than 10 mg/L) and produce clarified water with a turbidity of less than 2 NTU, suitable for pretreatment filters. However, in the case of highly turbid sea water, a conventional sedimentation system may not produce water with the desired turbidity level. To overcome this limitation, conventional sedimentation systems are often coupled with lamella plates ^[1][63]. These inclined plates increase the effective surface area for settlement, which allows for a smaller system footprint compared to conventional tanks. Studies have shown that the addition of lamella plates can increase sedimentation efficiency by up to 20% ^[34][114]. Lamella sedimentation tanks are often used to treat open ocean intake source water that is heavily influenced by river water or wastewater discharges with high turbidity ^[3][57]. A life cycle analysis (LCA) was performed to compare the energy consumption of two RO pretreatment methods, one using sedimentation-based pretreatment at the Fujeira 1 Desalination plant in the UAE, and the other using simulated membrane pretreatment ^[31][111]. The results showed a significant difference in energy consumption between the two systems, with the membrane-based pretreatment system being less energy-intensive. However, it is worth noting that cleaning the membrane-based system still requires more energy than sedimentation-based pretreatment. Additionally, it is important to note that the study compared a real sedimentation system to an idealized membrane pretreatment system and thus, simulated results should be followed by pilot studies as real systems may have higher energy consumption. It is critical to note that the sedimentation process has the drawback of not being able to remove all suspended solids, such as algae, which can lead to short filter runs, clogging of the subsequent granular media filters, and cause biological fouling on the RO membrane ^[19][99].

DAF is a water treatment method that is an alternative to conventional sedimentation. It is frequently used to treat water when suspended particles cannot easily settle. DAF can remove high concentrations of suspended solids, up to 8000 mg/L, and is particularly effective at removing low-density particles such as algae and natural organic matter that cannot be eliminated by conventional sedimentation ^[35][36][115,116]. The process involves creating fine air bubbles that attach to flocs and suspended matter, causing them to float to the surface. The floating flocs are then skimmed off, and clarified water with low turbidity is collected from the bottom of the tank ^[1][63]. The performance of a DAF unit depends on the air-to-solids ratio. DAF has mostly been examined as a pretreatment method for industrial waste waters ^[37][117], and more recently for seawater applications. Coagulation combined with DAF and in situ-generated liquid ferrate or ferric chloride has been proposed as a strategy for removing algal cells and organic matter from seawater ^[38][118]. However, this should be followed by pilot and cost studies, and the removal of residual iron, which can cause corrosion, must also be considered. A full-scale DAF process was investigated in a drinking water treatment plant ^[19][99]. It was found that the integration of DAF with the pre-sedimentation stage could be an effective method to produce stable water quality, as the pre-sedimentation process is crucial in removing heavy particles that may damage the DAF process. Alayande et al. ^[36][116] have recently conducted a comprehensive review of the current methods for controlling fouling in SWRO desalination plants and found that DAF pretreatment processes were employed in 16 out of 22 plants in the Arabian Gulf and Gulf of Oman. For example, the Fujairah II seawater RO desalination plant in the UAE incorporated the DAF system in combination with granular media filtration as a successful pretreatment technology for harmful algal blooms. However, the DAF process is still limited in its ability to remove nano- and pico-algae, as well as extracellular toxins, which are prevalent in some seawater ^[39][40][119,120].

Granular media filtration is a common pretreatment technology in existing full-scale SWRO plants ^[2][86]. Suspended or colloidal particles in water, remaining after the clarification process, may be effectively removed through media filtration. When water passes through a bed of porous and granular material, contaminants in the water are captured by the medium, leaving only clear water ^[41][121]. The filtration process involves the mechanisms of diffusion, interception, inertial compaction, adsorption, staining, and sedimentation. The performance of seawater pretreatment filters is influenced by the type, uniformity, and size of the filter medium, as well as the geometry of the contaminant particles, as highlighted in reviews by Anis et al. ^[19][99] and Jacangelo et al. ^[2][86]. It has been shown that granular media filters can reduce turbidity and improve water clarity by removing particles as small as 10 μm ^[19][99]. These filters use materials such as sand, gravel, garnet ^[42][122], magnetite, anthracite, and activated carbon ^[43][123], and often employ a combination of two or more materials in layers, such as in dual-media filtration (DMF). The main benefits of DMF are a high filtration rate, long runs, and a low RO feed-water silt index ^[19][99]. Studies have also found that using anthracite and sand together in filtration provides all the benefits of single-media filtration but requires less backwash water and allows for higher filter rates than using either one alone ^[44][124]. Research has shown that using a layer of granular activated carbon in dual-media filters is effective in removing high levels of organics. This method has the benefit of organics removal by adsorption, resulting in cost savings while reducing chlorine demand and producing safer water ^[45][46][125,126]. It can even remove remaining free chlorine from the chlorination pretreatment step ^[3][57]. Recent studies ^[47][127] have also examined the use of calcite ooids, a novel filter bed created through the seawater softening process, as a pretreatment stage in seawater desalination plants. The use of this new adsorbent media was found to remove 89.4% of turbidity and 66% of total organic carbons, with the potential to reach 95.7% removal after a granular activated carbon filter. Conventional seawater pretreatment filters can be either gravity or pressure-driven. These filters are used in desalination plants and operate in a down-flow manner. Gravity-driven filtration, which uses open-atmosphere filtration beds, is a cost-effective option for large desalination plants. However, pressurized media filtration is also widely used in small desalination plants, as it requires less space and can be installed faster. For example, the pretreatment in the Fujairah II SWRO plant is based mainly on coagulation and dual-media gravity filters, producing a water quality of SDI 2.7 ^[21][101]. Pressurized media filtration has also been widely used in small desalination plants since they require less space and can be installed faster ^[19][99]. However, additional studies are needed to reduce the energy cost of these pressure filters. Cartridge filters, which use 1–10 μm filters, are often used as a last pretreatment stage to remove remaining suspended solids ^[48][128]. However, they are mainly used as a protection device and do not perform significant silt removal. A well-designed granular media filtration system can improve the performance of an RO membrane, but organic and biofouling are still limiting factors ^[49][129]. Modifications to the design and operational parameters of media filtration are needed to alleviate the organic and biofouling of the RO membrane.

Scaling occurs on the surface of the RO membranes because of salt and mineral precipitation from seawater. This precipitation, which often occurs in the final stage of installation, is caused by supersaturation. Scaling reduces permeate flow and shortens the lifetime of membranes. Scale inhibitors, also known as antiscalants, can be used as a conventional pretreatment technique to control scaling by introducing them at concentrations typically below 10 mg/L ^[50][130]. These inhibitors work by adsorbing or reacting with the active growth sites of the scale matrix ^[51][131]. The use of antiscalants depends on the water quality, concentrate discharge limits, and targeted recovery. It has been shown that scale inhibitors are recommended when desalination plants operate at a recovery rate of 35% or higher ^[2][86]. Various chemicals, such as sodium hexametaphosphate, organophosphates, polyphosphates, and polyacrylates, can act as antiscalants ^[10][92]. However, the correct dosage of these chemicals can be difficult to calculate and may require the use of a simulation tool. Overdosing can lead to biofouling of the RO membrane ^[52][53][44,132]. In fact, these chemicals can act as a nutrient for bacteria, leading to their attachment to the membrane. Additionally, the presence of phosphorus-based scale inhibitors in brine discharge may result in algal blooms around the discharge area ^[5][49]. Another issue is that residual cationic flocculants from the pretreatment process may react with some antiscalants to create sticky foulants ^[19][99]. Some desalination plants control scaling by adding acid, such as sulfuric or hydrochloric acid ^[52][44]. For economic and safety reasons, sulfuric acid is the most used. However, research has shown that it may increase the risk of sulfate scaling, such as barium sulfate scaling ^[19][99]. It should also be noted that the use of acid can cause corrosion of equipment and shorten the lifetime of membranes ^[11][60]. Anis et al. ^[19][99] have analyzed the toughest challenges to overcoming the drawbacks of conventional scale inhibitors, such as the utilization of environmentally friendly and non-phosphorus-based scale inhibitors with the optimization of operating procedures. As a result, various alternative green inhibitors against scale formation have recently been suggested ^[50][54][130,133]. However, most of the research is limited to laboratory-scale testing.

Membranes synthesized from polymeric materials are generally organic in nature. These membranes are comparably more cost-effective than those manufactured from inorganic materials or ceramics ^[55][143]. They are easily handled during the fabrication process and can be made into a variety of different configurations for optimum performance, typically with a high water production capacity ^[55][56][57][143,144,145]. The permeate quality and the operating cost of water production are greatly influenced by the type of polymer material used during the fabrication process. Therefore, it is crucial that the most appropriate polymer and pore size are properly selected for any filtration process to avoid long-term problems such as frequent membrane replacement and high energy consumption.

MF membranes are large pore-sized membranes that range from 5 µm to 0.1 µm with the ability to filter out emulsions of latex, blood particles, cells, and bacteria ^[10][58][92,139]. In the last 15 years, extensive research has been conducted to establish these MF membranes and determine their validity as an alternative pretreatment method to conventional techniques. it was found that MF membranes were able to increase the water flux and reduce the SDI ^[59][146] without requiring changes in the feed water pH, and were effective in microbial elimination ^[1][60][63,147]. Nevertheless, one of the major disadvantages of MF is membrane fouling. Foulants including proteins, organic algae, and oil particles are the main contributors to MF membrane fouling ^[61][140]. Although membrane fouling is the most significant disadvantage of MF membranes, low thermal stability and resistivity toward free chlorine are other limitations ^[62][63][64][148,149,150]. Many studies have emphasized the importance of modifying these membranes for better performance and utilizing hybrid pretreatment techniques to overcome this issue, which are going to be discussed later ^[65][151].

UF membranes, based on multiple studies, were found to be more efficient than MF membranes and the conventional pretreatment methods. UF membranes are cost-effective with a better removal capability of silt, suspended organics, and microbes, resulting in a high and consistent filtrate quality with low RO fouling potential ^[3][66][67][57,152,153]. SWRO systems designed with UF membrane pretreatment are often termed “dual-membrane systems” ^[68][154]. In Wang Tan Power Plant, a dual-membrane system was installed and operated with low flux and low chemical treatments and resulted in a high filtrate quality with SDI < 2.5 and a reduction in turbidity of 98–99.5% ^[69][155]. In Singapore, experimental studies showed that using sand filtration, MF membrane, and UF membrane resulted in a filtrate quality with SDI between 2.8 and 6.3, 2.0 and 3.0, and 1.0 and 2.0, respectively, showing the superior efficiency of UF membrane technology compared with MF and conventional methods ^[70][156]. However, as in MF membranes, UF suffers from membrane fouling that affects its performance. The primary foulants are referred to as natural and effluent organic matter (NOM and EfOM) ^[71][72][73][157,158,159]. Modifying the UF membrane, using for example graphene oxide nanosheets ^[74][160] and/or integrating it with other conventional pretreatment methods, such as coagulation, has been identified by Yang et al. ^[75][161] as a viable solution to solve the UF fouling problem. The hybrid system showed good permeate quality compared to stand-alone pretreatment units.

While NF, a pressure-driven membrane process, is considered a promising membrane pretreatment method and represents a major milestone in the membrane technology field as it has a high retention capability of divalent ions and salts ^[76][77][78][162,163,164]. NF membranes, with a typical pore size of 1 nm, operate as porous and non-porous membranes as they transport in sieving and diffusion mechanisms ^[79][80][165,166]. These membranes are typically capable of achieving the elimination of divalent and monovalent ions in the range of 75–99% and 30–50%, respectively ^[81][167]. They are capable of eliminating microorganisms, turbidity, and a part of the dissolved salts ^[82][168]. Furthermore, they can significantly and efficaciously reduce the scaling of the RO membrane by eliminating Ca²⁺, Mg²⁺, and SO4²⁻ ions ^[83][169] with a reduction in total hardness of 86.5% ^[82][168]. In the Umm Lujj SWRO plant, Saudi Arabia, NF combined with their system resulted in a noticeable growth in the permeate flow rate from 91.8 to 130 m³/h ^[84][170]. Therefore, extensive research studies have been carried out on NF membrane and considered it as a confirmatory step to be applied in RO pretreatment systems ^[82][83][168,169]. However, as in MF and UF membranes, NF membranes encounter the same fouling issue that results in an energy consumption increase and a reduction in the lifetime of the membranes ^[81][85][167,171]. Inorganic foulants including metal hydroxides, carbonate, and sulfate-based salts are among the common NF foulants that cause membrane scaling ^[86][172]. Therefore, fabricating modified NF membranes with anti-scaling properties is of great value to maintain the performance of the NF technology ^[87][173].

For decades, numerous fabrication techniques have been employed in the preparation of polymeric membranes for a wide range of applications. The choice of fabrication technique depends on the polymer and the desired membrane structure. Non-solvent-induced phase separation (NIPS), evaporation-induced phase separation (EIPS), thermally induced phase separation (TIPS) and vapor-induced phase separation (VIPS), and other fabrication methods including interfacial polymerization, stretching, track-etching and electrospinning are among the commonly used techniques for the preparation of polymeric membranes ^[88][181]. These techniques aid in tailoring the membranes’ morphology, mechanical properties, pore-size distribution, hydrophilicity, selectivity, fouling mitigation, and flux ^[89][182].

Many studies have reported improvements in the morphology, properties, and performance of the prepared membranes simply through the addition of inorganic and high molecular weight organic materials such as polyvinylpyrrolidone (PVP), poly(ethylene glycol) (PEG), Arabic Gum, or lithium chloride (LiCl) as additives to the polymer solution ^[90][91][183,184]. The addition of LiCl increases the viscosity of the casting solution due to its strong interactions with the polymer and solvent, thus improving the membrane’s permeability and rejection ^[92][93][94][185,186,187]. Similarly, the incorporation of PVP enhances the membranes’ performance owing to the increased hydrophilicity and pore density. PVP causes a decrease in the effective thickness of the dense layer due to the formation of macro-voids in the support layer ^[95][188]. Other filler materials including metal/metal oxide or carbonaceous nanoparticles have also been reported to enhance the anti-fouling characteristics and permeate flux of membranes. For instance, a recent study reported the fabrication of oxidized carbide-derived carbon-incorporated polyether sulfone UF membranes prepared via NIPS ^[96][189]. The prepared membranes demonstrated improved porosity, pore size, and surface free energy with a noticeable reduction in the water contact angle. The membranes revealed a humic acid (HA) rejection of 96.8% and a maximum flux recovery ratio (FRR) higher than 86.7% over three cycles of pure water/HA filtration over a period of 140 min, suggesting excellent stability and reusability of the membranes.

Although the fabrication technique has a significant effect on the performance of the polymeric membranes’ separation, there are other factors that highly influence the membrane performance including the membrane’s hydrophobic/hydrophilic nature, surface charge, porous structure, and surface roughness ^[97][196]. Materials used for fabricating hydrophobic membranes include polypropylene (PP), PES, polytetrafluoroethylene (PTFE), and PVDF ^[98][178]. However, hydrophobic membranes were reported to be more prone to membrane scaling compared to hydrophilic membranes, which significantly affects their performance. Hence, various attempts were conducted to improve the hydrophilic nature of the membranes through modification using, e.g., polyethylene glycol (PEG), polyvinyl pyrrolidone (PVP), guar gum, and cellulose in the membrane casting solution ^[99][197]. For instance, Cheng et al. ^[100][198] used tunicate cellulose nanocrystals (TCNCs) in membrane fabrication, which resulted in a super-hydrophilic membrane surface ^[100][198]. The surface charge of the membrane is another parameter that is considered while modifying and fabricating polymeric membranes. Chen et al. ^[101][199] prepared Iron (II) phthalocyanine (FePc)/PVDF membrane for oil treatment, which resulted in 96.7% oil rejection with 158.94 L/m²h water flux, where Fe played a great role in increasing the negative charge of the membrane surface. Liu et al.’s ^[102][200] study revealed that among various carbon-nanotube (CNT)-modified polymeric membranes for oil filtration, the PVA/CNT membrane presented a superior anti-fouling property owing to its hydrophilicity and the carboxyl functional groups that increased the surface negative charge where the flux recovery ratio (FRR) was 100%. A negatively charged membrane is less susceptible to membrane fouling due to electrostatic repulsion towards oil droplets and other foulants such as bacteria cells ^[103][104][201,202]. Therefore, most of the recent research has been directed towards modifying the membrane surface by increasing its surface negative charge ^[105][203].

Membrane pore structure is one of the most crucial morphological criteria of polymeric membranes that influences their efficiency ^[106][204]. The larger the pore size of the membrane (>100 nm), the faster the irreversible fouling of the membrane compared with narrow pore membranes (30 nm) ^[107][205]. Surface roughness is a further membrane characteristic that has an influence on the membrane performance. Increasing surface roughness has been reported to increase permeability and diffusion due to the increase in the cross-sectional area ^[98][178]. However, high surface roughness is known to be a key factor that leads to membrane fouling. This is ascribed to the accumulation and absorption of the foulants onto the membrane surface, which consequently causes a reduction in water flux ^[108][206]. To avoid this, efforts have been concentrated on fabricating membranes with low surface roughness. An oil separation study conducted by Panda et al. ^[109][207] demonstrated that by varying PAN concentration in a PAN/PEG membrane, a higher concentration demonstrated a higher surface roughness (35 nm), which resulted in a higher fouling tendency compared with the low surface roughness membranes (10 nm). By reducing surface roughness, a dramatic reduction in the flux decline ratio (FDR) was reported from 55% to 25%. Modifying the membrane is a way to reduce membrane roughness. For example, Wan Ikhsan et al. ^[110][208] revealed a reduction in PES membrane surface roughness after the employment of halloysite nanotube-hydrous ferric dioxide nanocomposite (HNT-HFO), which induced a significant improvement in oil rejection of about 99.7%.

Despite the contribution of the aforementioned factors in the performance and application of polymeric membranes, the configuration of the membranes is an essential aspect that is to be considered during membrane fabrication and selection ^[98][178]. The most commonly developed module configurations are tubular, spiral wound, hollow fiber, and plate types ^[111][209]. Plate and tubular modules are among the highest-cost and lowest-industrially practical configurations. This is because plate membranes are more prone to fouling ^[112][210] while tubular membranes are energy-intensive ^[113][211]. Hence, due to the problems generated by the plate and tabular membrane configurations, most water treatment plants have replaced them with spiral wound and capillary fiber membranes ^[98][178]. Although spiral wound membranes are energy-intensive, various features have been reported compared with other membrane configurations, presenting their high packing density, high salt rejection, simple construction and operation, less cleaning frequency, and effective flow mixing ^[114][212]. Capillary fiber membranes are selective membranes with great packing density and strong resistance to filtration pressure ^[115][213]. This makes both configurations the selected types of membrane in large water treatment plants.

As mentioned above, polymeric and ceramic membranes are the most used materials for filtration and separation processes due to their high performance. However, the main drawbacks of using polymeric membranes are fouling, biofouling ^[116][214], and sensitivity to pH and temperature ^[117][215], which can decrease flux and selectivity. As a result, many efforts have been made to improve these properties in polymeric membranes or find alternatives. Accordingly, a recent increase in research on ceramic membranes for UF, MF, and NF processes as alternatives to polymeric membranes was observed. Therefore, a comprehensive literature search was carried out for all studies related to performance factors on both ceramic and polymeric membranes. While ceramic and polymeric membranes are used in various water pretreatment applications, ceramic membranes present competitive advantages over polymeric membranes. The excellent chemical resistance achieved by ceramic membranes make them strongly competitive against other commercial membranes as they are able to withstand a wide range of chemicals, including strong acids and bases ^[118][216], contrary to the polymeric membranes that might be sensitive to certain strong acids and bases ^[119][217]. Ceramic membranes also have a higher temperature tolerance and can operate at high temperatures of up to 500 °C without degrading. This can be beneficial in removing irreversible foulants from the ceramic membrane surface. In comparison, fouling and biofouling are among the biggest challenges in applying polymeric membranes. In this respect, Sarkar et al. have recently presented a comprehensive review of polymeric membranes and highlighted that this type of membrane has some limitations, including poor thermal and mechanical stability, sensitivity to salinity, and lower lifetime ^[120][218]. While ceramic membranes have a higher initial cost due to the use of expensive materials and manufacturing processes, they also have a longer lifetime and can operate for several years before needing replacement ^[118][216]. This can reduce the frequency and cost of maintenance and membrane replacement, resulting in lower overall costs over the membrane’s lifespan. On the other hand, it should be noted that some ceramic membranes use unrefined raw materials such as clays, zeolites, apatite, fly ash, and rice husk ash ^[121][219]. These low-cost raw materials could offer a promising approach to reducing the capital cost of ceramic membranes. Li et al. and Kommineni et al. have reviewed the advantages of ceramic membranes, including fouling resistance, high permeability, good recoverability, chemical stability, higher mechanical robustness, ability to handle higher loading of particulates, and long lifetime ^[118][122][216,220]. It seems that ceramic membrane pretreatment can be a cost-competitive option and a critical player in water technology ^[123][221]. Additionally, hybrid membranes, which combine ceramic and polymer materials, can also be used to improve membrane performance. In the following section, these types of membranes are discussed in more detail.

It is important to note that these provided values are just approximate averages, and actual results may vary depending on the specific operating conditions, type of feed water, and type and size of the ceramic or polymeric membrane.

In the last two decades, the use of ceramic membranes (CM) for water desalination has attracted significant attention because of their excellent properties such as high flux rates; reliability; thermal, mechanical, and chemical stability; ability to withstand harsh environments such as acidic conditions; and ease of cleaning ^[124][125][225,226]. Despite the fact that the investment cost of CM is higher than polymeric membranes, the overall cost can be compensated by longer lifetimes and higher permeate fluxes ^[125][226]. Moreover, the pore size of CM can be tuned during the manufacturing process toward specific applications. According to the application of CM, different fabrication techniques including pressing, slip casting, and extrusion were utilized to produce micro, ultra, and nanofiltration CMs ^[19][99]. Moreover, the geometry and configuration of CM (flat sheet, hollow fiber, or tubular) can be produced based on the support and the required application ^[124][225].

Alumina, zirconia, zeolite, natural clays, silica, and titania are the most widely used materials for CM manufacturing. Similar to polymeric membranes, the quality of the feed water and the desired quality of the permeate flux inform the selection of the pore size of CM ^[125][126][226,227]. Moreover, the active surface of CM can be altered based on the quality of the feed water and toward a specific pore size of the membranes. For instance, Nogochi et al. ^[127][228] investigated seawater treatment by a commercial flat sheet CM with a pore size of 0.1 µm. The results demonstrated that the permeate flux from the CM has a turbidity of 0.04–0.1 NTU and SDI of 1.6–2.2, indicating that the system can provide high-quality water for RO treatment.

Cui et al. ^[128][229] utilized the ZrO₂/Al₂O₃ CM in a seawater desalination pilot plant in Tianjin Bohai, China. The raw seawater was pretreated by different methods, and it was concluded that flocculation and natural sedimentation was the optimum method for pre-CM filtration. The CM was used in two configurations, honeycomb and multichannel. The study concluded that, for the seawater obtained in Tianjin, China, coagulation is required before CM filtration. The filtration results demonstrated a high removal of turbidity, and the permeate SDI was 0.18–1.1. The membrane maintained a stable permeability for a long time even at low temperatures (3–6 °C). Achiou et al. ^[129][230] fabricated a tubular CM from natural pozzolan for the treatment of raw seawater. The CM was prepared by the extrusion and sintering method at 950 °C.

Two-layer CM was obtained by mixing natural clay and pozzolan for the support layer, and pozzolan powder was deposed on the inner surface of the tube by the crossflow filtration technique as a filtration layer. The fabricated membrane demonstrated an average pore size of 0.37 microns and a porosity of 41.2%. Filtration tests showed rejection of 98.3% and 70.8% for turbidity and COD for an initial concentration of 6.29 NTU and 5.69 ppm, respectively.

In another study, seawater pretreatment was conducted by a zirconium-based ceramic membrane with 0.05 µm pore diameter, and the obtained results demonstrated that the sea water flux and COD rejection rate were significantly enhanced due to the disruption of the adsorption layer and membrane surface fouling layers with heavy flow rates ^[130][231]. Wang et al. ^[131][232] synthesized a hallow fiber γ-aluminum coated on α-aluminum ceramic by changing the aluminum nanoparticle solution soaking time to investigate the influence of layer thickness on flux rate. The obtained results revealed that pure water permeate flux significantly increased with the mean pore size of 1.61 nm. In addition, the high multi-valent cation rejection rate was increased compared to monovalent cations since the aluminum-coated ceramic membrane demonstrated a positive surface change nature.

De Friend et al. ^[131][232] modified the aluminum-based ceramic membrane with different metals including Fe, Mn, and La. The filtration results showed that the water permeate flux was 50% higher compared to Mn-modified and bare aluminum ceramic membranes, which might be due to the chemical properties of modified aluminum ceramic membranes.

In another study, Belgada et al. ^[132][233] studied the effect of sintering temperature on the characteristics of raw phosphate CM. At the optimized sintering temperature (1000 °C), the membrane showed an enhanced permeability flux of 697 L/(h∙m²∙bar) and porosity of 25.6%. The CM tested for raw seawater treatment, and the results revealed a promising rejection of turbidity and TOC of 98% and 73%, respectively. The membrane has shown a 40% reduction in the SDI, in which the authors concluded that the seawater permeate product has sufficient quality for RO treatment. According to the flux recovery and fouling analysis results, both intermediate pore blocking and cake layer formation fouling mechanisms were identified; however, the membrane revealed 74.3% of seawater flux recovery after cleaning. Cui et al. ^[133][234] reported the influence of crossflow velocity on the flux of commercial CM with a pore size of 50 nm. It was found that the crossflow velocity has a great effect on the water permeate flux in laminar and turbulent regions. The permeate coming out from the CM has a turbidity of less than 0.1 NTU. Bottino et al. ^[134][235] treated lake water from Genoa, Italy by alumina CM with an average pore size of 200 nm. A complete removal of algae and microorganisms was achieved by the CM. Moreover, rejection rates of 56% and 64% were obtained for chloroform and TOC, respectively.

Kang et al. ^[135][237] tested alumina commercial CM (pore size 100 nm) for synthetic seawater treatment. The study concluded that the best removal of turbidity (0.076 NTU) and lowest value of SDI (0.9) were achieved utilizing 6 mg/L of FeCl₃ as a coagulant. Moreover, the permeate flux and DOC removal were significantly improved by coupling coagulation and CM filtration. Islam et al. ^[136][238] synthesized porous supported YSZ (Yttria Stabilized Zirconia) CM using an atmospheric plasma spraying technology. This technique yields a high porosity with a lower pore size of YSZ membrane, and the YSZ membranes are homogenous and defect-free in nature. The YSZ CM exhibited remarkable filtration results for permeate flux, rejection rate, and permeability for three different contaminated water sources including waste and salt water up to 400 Lm⁻² h⁻¹, ~95%, and 380 Lm⁻² h⁻¹ bar⁻¹, respectively. In addition, the YSZ membrane also presented excellent cycling stability. Based on the filtration results, the authors concluded that YSZ-coated ceramic membranes showed better performance than commercial ceramic membranes.

Dong et al. ^[137][240] developed a novel technique to synthesize thin-film nanocomposite nanofiltration membrane. The TFC was coated with zeolite nanoparticles, and then a polyamide layer was formed on the zeolite surface layer using interfacial polymerization. The characterization results revealed that zeolite nanoparticles enhanced the membrane surface roughness and resulted in high permeability compared to bare TFC. CM membranes have great potential for the RO pretreatment process; however, further research is needed to fully understand their impact on RO pretreatment processes.

Although significant membrane enhancement has been achieved, the industrial application of ceramic membrane is still hindered due to the fouling problems, which results in high operating costs and huge capital investment because of the periodic membrane cleaning and replacement. Several approaches have been devoted to mitigating the fouling phenomenon, including membrane cleaning, membrane module design, and modification of membrane surfaces ^[138][139][241,242]. For instance, Xu et al. ^[140][243] applied different membrane cleaning processes including chemical cleaning, backwash, and ferric coagulation to improve the applicability of commercial UF membranes as a pretreatment of RO seawater desalination.

Hybrid pretreatment systems can be defined as the combination of one or more conventional pretreatment units with one or more of membrane pretreatment (MF, UF, and NF). These systems are a viable and efficient option as they utilize the strength of different units. Additionally, due to severe saltwater conditions, which increase the risk of membrane fouling, these systems are often used in commercial SWRO plants. Conventional pretreatment methods such as DAF, coagulation, and chlorination are used to provide a contaminant barrier before the water reaches membrane units ^[19][99].

Starting with MF membranes, Ebrahim et al. ^[141][249] conducted a study in Kuwait that demonstrated RO pretreatment design where the feed water is chemically treated with chlorine and filtered by a coarse strainer before being fed to the MF membrane. With this design, an excellent SDI value of the filtrate was observed (2.22% average value). As biofouling is considered a major industrial problem, a hybrid chlorination–MF system was evaluated by Lee et al. ^[142][250].

UF is considered the most used membrane pretreatment in hybrid systems. Glueckstern et al. ^[143][252] tested the UF membrane performance in an SWRO system with a hybrid pretreatment method. Screen filtration, coagulation, and chlorination were applied prior to the UF membrane. This resulted in a good filtrate quality where SDI and turbidity ranged from 0.8 to 3.8 and 0.1 to 0.2 NTU, respectively. Villacorte et al. ^[49][129] reported that combining coagulation with UF membrane technology can reduce the fouling potential that is caused by harmful algal blooms. These results support the experimental study conducted by Kim et al. ^[144][253], who applied coagulation/flocculation before the DMF and UF membrane. The results showed the SDI value was 6.0 and 2.0, respectively. The Heemskerk water treatment plant, which is located in the Netherlands, utilizes integration between coagulation-sedimentation filtration and a UF system prior to RO ^[145][254]. The results showed that this integration results in superior particle elimination, which resulted in the mitigation of colloidal fouling. In addition, a recent study carried out by Monnot et al. ^[146][255] demonstrated the feasibility of utilizing granular activated carbon (GAC) pretreatment before UF to reduce its fouling potential and increase efficiency in the removal of dissolved organic carbon (DOC). In addition to GAC, powdered activated carbon (PAC) with UF were combined as a pretreatment to SWRO by Tansakul et al. ^[147][256]. The addition of PAC in the UF process enhanced the performance of UF; the UF fouling rate was reduced, and the NOM retention rate increased from 10% to 45% without and with PAC, respectively. In the same study, Tansakul et al. ^[147][256] studied the effect of utilizing a low-cost and widely available bentonite adsorbent as a conventional pretreatment to UF. However, this addition has no significant effect on UF performance. Park et al. ^[148][257] studied the combination of DAF technology with a membrane-based filtration system. The Al-Shuwaikh desalination plant in Kuwait equipped with DAF/UF systems presented SDI < 2.5 for good quality feed water and <3.5 ^[148][257]. Yang and Kim ^[75][161] studied the effect of coagulation on the performance of MF and UF for the removal of particles with two types of membranes. The results showed that the SDI₁₅ of permeate from coagulation–MF and coagulation–UF were 0.75 and 1.88, respectively. While SDI₁₅ for only MF membrane permeate was 3.17 and 2.76 for UF only.

Numerous research studies have been conducted to assess the feasibility and efficiency of hybrid systems that couple NF membranes with conventional pretreatment ^[149][258]. Using NF as a pretreatment in SWRO not only improves the feed-water quality, for example, through the removal of hardness, turbidity, or microorganisms but also improves the entire desalination process ^[150][259]. NF membrane reduces the ionic salts content present in seawater, resulting in significantly reducing the osmotic pressure, and hence the RO unit can be operated at a lower pressure and subsequently with less energy along with a higher recovery rate ^[151][260]. For example, at 40 bars, the permeate flow and recovery from the conventional SWRO is only l l/m and 16.7%, respectively, as compared to a much higher flow of 4.8 l/m and recovery ratio of 48% using the new NF-SWRO process ^[151][260]. Park et al. ^[152][174] conducted a study aimed to minimize scale formation potential in RO membranes. They used a UF/NF/RO hybrid pilot system as a pretreatment unit to remove divalent ions from seawater. The results showed that the UF did not reject any ions because of pore size. The rejection of divalent ions by NF was in order of sulfate (>95%), magnesium (>60%), and calcium (>30%) in every rejection experiment based on a water recovery rate of (40, 50, 60, 70, and 80%). In the UF/NF/RO hybrid system, most of the divalent (>99%) and the monovalent (>97%) ions were effectively rejected with slightly increased divalent ion rejection compared to the UF/RO system ^[152][174]. In Saudi Arabia, fine and thick sand filtration media were utilized prior to NF and highly improved the feed-water quality ^[84][170]. In addition, the utilization of NF as a pretreatment in the SWRO desalination pilot plant enhanced the production of water by more than 60%, which led to a cost reduction of 30% ^[153][261].

In addition to the polymeric membranes (MF, UF, and NF) discussed in the previous section, ceramic membranes have been recently employed in many applications ^[154][155][267,268], to replace conventional methods ^[156][269], as previously discussed. Hybrid ceramic technologies have gained attention in recent years as a potential solution for various applications. These technologies combine the advantages of ceramics with those of other technologies such as conventional methods. As mentioned in the previous section, in the study by Cui et al. ^[128][229], a hybrid ceramic membrane was tested in a seawater desalination pilot plant using different configurations, as pretreatment to SWRO. Additionally, the effectiveness of using a hybrid ceramic adsorption filter (CAF) and UF pretreatment in reducing RO fouling was studied by Nakano et al. ^[157][270]. The results showed that the CAF could eliminate a fraction of the dissolved organic matter that escaped from the UF membrane, thus reducing RO membrane fouling. In addition, the results showed that the use of CAF pretreatment reduces the amount of biofilm formed on the RO membrane, which delays the reduction in RO membrane permeability and the membrane cleaning frequency. These two factors could reduce the operating costs of seawater desalination plants and enhance capacity utilization, leading to lower costs of water production ^[157][270].

The water cost is considered one of the important factors in the selection of water desalination and pretreatment technologies. The pretreatment technology’s economic analysis based on total water cost indicates that the membrane pretreatment system is less expensive than the conventional pretreatment system by 3–4% ^[1][63]. The total water cost for facilities using a conventional system is USD 0.59/m³ and for facilities using membrane systems, it is USD 0.55/m³. However, when considering capital costs, membrane systems prove to be more expensive than conventional systems by 2%. The RO membrane requires two cleanings per year whereas the conventional systems require nine cleanings per year, which dramatically contribute to the costs of the plant operation ^[1][63]. Hybrid technologies have the potential to reduce costs as they can combine the advantages of different systems while minimizing their drawbacks. However, it is important to note that the cost of hybrid pretreatment depends on the technologies employed and other factors such as local regulations. Unfortunately, the water cost using a hybrid pretreatment system has not been well studied, and further investigation on this aspect is a must to understand the cost-effectiveness of the hybrid approach.