Seawater desalination is a globally expanding coastal industry with an installed capacity of 80 million m3/day as of 2013. Reverse osmosis (RO) has become the dominant technology for seawater desalination with more than two thirds of the global installed desalination capacity. The major challenge for cost-effective application of seawater RO (SWRO) systems is membrane fouling. To mitigate fouling and reduce associated operational problems, pretreatment by granular media filtration (GMF) or micro- and ultrafiltration (MF/UF) is commonly required. Operation of SWRO pretreatment has proven to be challenging during algal bloom periods where relatively high concentrations of algal cells and algal organic matter (AOM) are present in seawater. Experience from a severe red tide bloom in the Middle East in 2008-2009 showed that GMF in combination with coagulation cannot handle severe algal bloom events. During this period, the failure of GMF to produce acceptable RO feed water quality (silt density index, SDI < 5) caused the shutdown of several desalination plants in the region. This event highlighted the importance of reliable pretreatment systems for SWRO operation, and focused the attention of the desalination industry on MF/UF technology. MF/UF systems are generally more reliable than GMF in producing stable, high quality RO feed water in terms of turbidity and SDI. Moreover, MF/UF product water quality is not affected by variations in raw water quality. Experience with large-scale UF operation in SWRO pretreatment during severe algal bloom events is limited and data is scarce. However, a well documented case of UF/RO operation on North Sea water in the Netherlands showed that during severe algal bloom periods, coagulation was required to stabilize UF hydraulic performance. In general, MF/UF membranes do not rely on coagulation to reduce turbidity and SDI. However, coagulation may enhance AOM removal in MF/UF systems and reduce particulate/organic and biofouling potential of UF permeate. From an operational point of view, it is desirable to completely eliminate coagulation from the process scheme, to reduce costs and complexities associated with chemicals, waste treatment, handling and discharge. The goal of this study was to evaluate the feasibility of UF as pretreatment to SWRO during algal bloom periods and to investigate the role of coagulation in improving UF operation. Ultimately the study aimed at providing insight into options for minimizing and ideally eliminating coagulation from UF pretreatment to SWRO. Algal blooms adversely affect UF operation by causing higher pressure development during filtration; lower permeability recovery after backwashing; and high concentration of algal biopolymers in UF permeate. The latter results in higher particulate/organic and biofouling potential of SWRO feed water. The first step of the study was to understand particle properties that affect fouling in MF/UF systems. Theoretical calculations indicated that spherical particles as small as a few nanometres - forming cake/gel layers with porosity ranging from 0.4 to 0.99 - do not contribute significantly to pressure increase in MF/UF systems operated at constant flux, indicating that the creation of large aggregates by e.g., extended flocculation, is not required in these systems. Further investigations were made to study the effect of process conditions on inline coagulation with ferric chloride prior to MF/UF. Experimental results showed that extended flocculation was not required for inline coagulation prior to MF/UF systems treating surface water and proper selection of dose and pH was sufficient to optimize MF/UF operation in terms of fouling potential and permeate quality. Calculations indicated that high G-values and short residence times encountered prior to and within MF/UF elements in practice, seem to be sufficient to maintain low fouling potential and control non-backwashable fouling. The effect of coagulation on hydraulic performance and permeate quality of UF membranes fed with AOM solutions in synthetic seawater was investigated. AOM biopolymers had high fouling potential as measured by the Modified Fouling Index (MFI) and were very compressible. Filtration at higher flux exacerbated both fouling potential and compressibility of AOM. Coagulation substantially reduced fouling potential, compressibility and flux dependency of AOM, resulting in substantially lower pressure development in filtration tests at constant flux. Inline coagulation/UF was more effective than conventional coagulation followed by filtration (0.45 µm) in terms of biopolymer removal at low coagulant dose (~ 0.5 mg Fe(III)/L). The applicability of coating UF membranes with a removable layer of particles at the start of each filtration cycle for treating algal bloom-impacted seawater was investigated. Iron hydroxide particles were applied as coating material at the start of each filtration cycle at different equivalent dose. Without coating, AOM filtration was characterized by poor backwashability. Pre-coating was effective in controlling non-backwashable fouling using ferric hydroxide prepared by simple precipitation and low intensity grinding. However, relatively high equivalent dose (~ 3 - 6 mg Fe(III)/L) was applied. Pre-coating with ferric hydroxide particles smaller than 1 µm - prepared by precipitation and high intensity grinding - resulted in low equivalent dose of 0.3 - 0.5 mg Fe(III)/L required for stable operation of the UF membranes. Further reducing particle size of the coating material is expected to be more effective in lowering the required equivalent dose. However, preparation of such particles requires further research efforts. Coagulation of AOM was studied for conventional coagulation (coagulation/flocculation and sedimentation) followed by filtration (0.45 µm), to identify AOM removal rates in seawater. Coagulation followed by sedimentation required coagulant dose of up to 20 mg Fe(III)/L to remove AOM biopolymers by up to 70%. Filtration through 0.45 µm had a significant impact on AOM removal, even at coagulant dose < 1 mg Fe(III)/L. This indicated that coagulated AOM aggregates have better filterability than settlability characteristics which may be attributed to the low density of these aggregates and could have considerable implications for the choice of clarification process in conventional pretreatment systems. The applicability of low molecular weight cut-off UF (10 kDa) membranes as a coagulant-free alternative to SWRO pretreatment was investigated. 10 kDa membranes were capable of completely removing AOM biopolymers from SWRO feed water without coagulation. UF membranes with nominal molecular weight cut-off of 150 kDa reduced biopolymer concentration to approximately 200 µg C/L (~ 60% removal). In terms of hydraulic operation, 10 kDa membranes showed lower permeability recovery after backwash than 150 kDa membranes. Physical characterization of the two membranes revealed much lower surface porosity of 10 kDa compared to 150 kDa membranes. In general terms, this study demonstrated that during algal blooms, UF membranes with nominal molecular weight cut-off of 150 kDa operated in inside-out mode, are more capable of reducing particulate/organic fouling potential of SWRO feed water at low coagulant dose than conventional coagulation. The application of UF membranes with low molecular weight cut-off can further enhance RO feed water quality in terms of particulate/organic fouling potential during algal blooms, without the need for coagulation. However, a small amount of coagulant may be required to control hydraulic operation of the UF membranes during these periods. Further improvements in material properties of these membranes should be directed at increasing the surface porosity of the membranes to enhance permeability recovery and ensure stable hydraulic operation.

Seawater desalination is a globally expanding coastal industry with an installed capacity of 80 million m3/day as of 2013. Reverse osmosis (RO) has become the dominant technology for seawater desalination with more than two thirds of the global installed desalination capacity. The major challenge...