Membrane Distillation (MD) utilizes thermally-induced vapor pressure difference between the feed and permeate as a driving force to allow only water vapor to pass through a hydrophobic microporous membrane. MD has distinctive advantages, such as tolerance to high salinity, moderate operational temperatures, pressures, and utilization of low-grade waste heat. Recently, MD has been gaining attention for its suitability for treating challenging industrial wastewaters and high salinity brines and implementing zero liquid discharge (ZLD).

Herein, first, we demonstrated an MD system with a Multi-Stage Flash Crystallizer (MSF-Cr) to achieve ZLD as a techno-economically feasible approach to extract valuable minerals from reverse osmosis (RO) brine. The proposed method separated Na₂SO₄ crystals from the feed, which was confirmed by an X-Ray Diffraction (XRD) analysis of the crystals extracted through a lab-scale experiment. With ASPEN Plus, various configurations with MD pretreated feed were simulated to determine impacts on the recovery rate of the distillate and salts. The gain output ratio grew with increasing numbers of stages from 2.77 to 4.0, stabilizing at a flow rate of 700 L/h at 70 ºC for 40 stages. A recovery rate reaching 4% for water and 25.05 kg/h for Glauber’s Salt (Na₂SO₄·10H₂O) was observed. The technoeconomic analysis revealed the optimal configuration to be a 7-stage module operating at 60 ºC and a flow rate of 700 L/h. Treatment costs were $1.17 per m³ of feed, which was reduced to $0.35 per m³ when beneficial resources were sold. Considering MD pretreatment, the combined treatment costs were $2.01 per m³ without the sale of valuable resources, which was highly competitive.

However, when inorganic salts, organic oils, and other low surface tension contaminants are present in the feed, there is a high chance for MD to suffer from membrane wetting, resulting in lower product water quality and system failure. Thus, recent MD research has been primarily concentrating on controlling and preventing wetting through optimizing feed pretreatment and operational conditions, applying new methods for detecting wetting, and novel membrane fabrication and surface modification. Thus, the second part of this research focused on Optical Coherence Tomography (OCT) enabled in-situ foulant monitoring in 3D. Although 2D imagery is widely applied in recent studies, this can only determine fouling layer thickness in severe fouling. Therefore, an advanced 3D imaging analysis technique using intensity range filters was proposed to quantify fouling layer formation from a single 3D image. This approach reduces computational power requirements and successfully separates the fouling layer from the membrane at the microscale. Moreover, this technique allows the evaluation of fouling layer thickness, fouling index, and fouling layer coverage in real-time. In the experiment, polyvinylidene fluoride (C-PVDF) and polytetrafluoroethylene (C-PTFE) membranes were used to treat a feed consisting of industrial textile wastewater. A 22 µm thick fouling layer composed of thin and disperse foulants were observed on the C-PTFE, which could not be observed using the 2D approach after 24 h. Moreover, the C-PTFE demonstrated better antifouling ability than the C-PVDF, as shown by its lower fouling index, supported by surface energy characterization. Discovering the potential of 3D imagery for fouling monitoring can lead to lowered operational costs and improved system stability.

Membrane wetting is one of the biggest obstacles impeding the practical application of MD, as it causes significant drops in operation efficiency and may even lead to system failure. The occurrence of wetting most frequently occurs due to surfactants in contaminated or low surface tension liquid. However, common monitoring approaches, such as conductivity and flux measurement, cannot explain the wetting phenomenon during the wetting process in detail. Therefore, the last part of this research examined how such damages can be prevented in MD operations by monitoring membrane wetting in real-time so that membrane wetting could be detected in the early stage. Previous studies have proposed impedance spectroscopy for early wetting detection, where an earlier and larger variation was observed with precise signal detection. This study proposed an analytical approach to estimate the wetting front, which is the average feed intrusion distance, based on the impedance value recorded in real-time operation. The proposed approach was used to quantify the mechanism of membrane wetting occurring a commercial polyvinylidene fluoride membrane in the presence of a surfactant, which provided more vital information that cannot obtained through conductivity and flux measurements.