This thesis examines the feasibility of Selectrodialysis as an innovative and efficient technology for nutrient recovery from greenhouse wastewater. The study aims to address the critical challenges of water quality degradation and Na+ accumulation in greenhouse operations, where the selective recovery of ions such as Na⁺, Ca²⁺, and Mg²⁺ is essential for maintaining optimal plant health and agricultural productivity.

The research evaluates the performance of SED by investigating the effects of various operational parameters, particularly voltage, and the presence of competing cations on ion transport efficiency. Experimental analysis revealed that SED could effectively achieve selective ion separation at an optimal voltage of 9V, which maximizes ion fractionation and migration rates while maintaining moderate energy consumption. However, higher voltages resulted in increased energy demands, water dissociation, and concentration polarization, which adversely impacted the selectivity of specific ions, particularly Na⁺.

The presence of competing cations, such as Ca²⁺ and Mg²⁺, was shown which significantly impacts the Na⁺ transport by reducing its migration rate and recovery efficiency. These effects are primarily due to competitive interactions and increased membrane resistance caused by concentration polarization, particularly at higher concentrations of divalent cations. Understanding these interactions is crucial for optimizing the recovery process and energy efficiency in systems where competing cations are prevalent.

The study also highlights the economic implications of implementing SED compared to conventional methods like RO. While SED demonstrates technical advantages in terms of selective ion recovery and reduced chemical usage, its broader adoption is constrained by significantly higher capital and operational costs. The total capital cost for SED was found to be over 13 times higher than that of RO, primarily due to the specialized membranes required for selective ion transport and the additional electrical components necessary for ion migration. Additionally, operational costs for SED are elevated due to higher energy consumption, which is more than double that of RO, and increased maintenance expenses resulting from the complexity of the system.

Despite its technical feasibility, the economic analysis suggests that SED may be better suited for niche applications where selective ion recovery is critical, such as in high-value greenhouse operations where precise nutrient control is essential. To improve the economic viability of SED, future research should focus on integrating renewable energy sources, experimenting with new membrane materials, and exploring different operational modes to reduce energy consumption and overall costs.

Organic micropollutants (OMPs) originate from organic chemicals such as drugs and pesticides that are widely used in human activities. OMPs are difficult to remove by conventional water treatment techniques, and hence continue to accumulate in natural water bodies. More effective methods need to be investigated for the removal of OMPs in drinking water treatment because of their toxicity and carcinogenicity, which may pose potential risks to human health. Previous studies have suggested that the use of activated peroxymonosulfate (PMS) catalyzed by Palladium (Pd) immobilized in ultrafiltration (UF) membranes can effectively degrade 1,4-dioxane and p-nitrophenol, while its removal efficiency for other OMPs, limiting factors and reaction mechanism still require for more research. In this study, PMS-Pd/UF system was established by coating Pd on the surface and 20 nm pores of the ultrafiltration membrane. The effectiveness of PMS-Pd/UF in the removal of OMPs from ultrapure water under various flux, pH, PMS dosages and ions presence was examined, as well as the performance in other water matrices including simulated brackish water, simulated brine water and river water. The results showed that PMS-Pd/UF achieved more than 95% of OMPs removal in 1 and 12-hour filtration tests at a flux of 30 LMH, while removal efficiency decreased with the increased flux due to the reduced contact time. The pH tests indicated that the system was more efficient under a neutral pH environment. The presence of 1 mM of various ions (Cl⁻, HCO3⁻, SO₄²⁻, and ClO⁻) had limited effects on the degradation of OMPs. However, the removal of OMPs was inhibited when OMPs coexisted with NOM in river water and, in combination with high salt concentrations simulated brine water (e.g., with 250mM of total anions). From specific scavenger dosing experiments, it was concluded that SO₄^•–, ^•OH, and O₂^•– were the main reactive species induced from Pd/UF-activated PMS for the removal of OMPs.

Large amounts of oily wastewater which included oil-in-water (O/W) emulsions, also known as produced water (PW), were produced in tandem with the enhanced oil recovery (EOR). Ultrafiltration (UF) was an effective and economic method to separate micron-sized O/W emulsions, while the membrane fouling limited its development. Understanding the fouling phenomena was essential to enhance the efficiency of membrane filtration for oil-water separation, hence this paper investigated the influence of different salinity (1 mM, 20 mM and 100 mM) and types of surfactants on the fouling of the alumina (Al2O3) membrane and the silicon carbide (SiC) deposited membrane filtrating 500 mg/L O/W emulsions with mean droplet sizes of approximately 4 ~ 7 μm, and the UF with the constant flux of 80 LMH and the crossflow velocity of 0.59 m/s was conducted. Sodium dodecyl sulfate (SDS, anionic), alkyl polyglycoside (APG, nonionic), and cetyltrimethylammonium bromide (CTAB, cationic) were chosen in this study due to the frequent use in EOR, and the N-dodecyl-N,N-dimethyl-3-ammonio-1-propanesulfonate (DDAPS, zwitterionic) was selected because of its low tendency to foul at high salt concentrations.
Derjaguin-Landau-Verwey-Overbeek (DLVO) and the extended DLVO (XDLVO) models were used to quantify the membrane-oil droplet and deposited oil layer-oil droplet surface interaction. The results showed that the SiC-deposited membrane had less membrane fouling and irreversible fouling resistance compared with the Al2O3 membrane when filtrating O/W emulsions stabilized with SDS, APG or DDAPS. The DLVO model estimated emulsion fouling propensity to rise with increasing salinity when dealing with SDS, APG or DDAPS-stabilized O/W emulsions, while CTAB-stabilized emulsion fouling propensity for the SiC-deposited membrane would decrease with the higher salinity.