This paper presents the development of a rigorous theoretical model to predict the transmembrane flux of a flat sheet hydrophobic composite membrane, comprising both an active layer of polytetrafluoroethylene and a scrim-backing support layer of polypropylene, in the direct contact membrane distillation (DCMD) process. An integrated model includes the mass, momentum, species and energy balances for both retentate and permeate flows, coupled with the mass transfer of water vapor through the composite membrane and the heat transfer across the membrane and through the boundary layers adjacent to the membrane surfaces. Experimental results and model predictions for permeate flux and performance ratio are compared and shown to be in good agreement. The permeate flux through the composite layer can be ignored in the consideration of mass transfer pathways at the composite membrane. The effect of the surface porosity and the thickness of active and support layers on the process performance of composite membrane has also been studied. Among these parameters, surface porosity is identified to be the main factor significantly influencing the permeate flux and performance ratio, while the relative influence of the surface porosity on the performance ratio is less than that on flux.

Fouling development in direct contact membrane distillation (DCMD) for seawater desalination was evaluated combining in-situ monitoring performed using optical coherence tomography (OCT) together with destructive techniques. The non-invasive monitoring with OCT provided a better understanding of the fouling mechanism by giving an appropriate sampling timing for the membrane autopsy. The on-line monitoring system allowed linking the flux trend with the structure of fouling deposited on the membrane surface. The water vapor flux trend was divided in three phases based on the deposition and formation of different foulants over time. The initial flux decline was due to the deposition of a 50-70 nm porous fouling layer consisting of a mixture of organic compounds and salts. Liquid chromatography with organic carbon detection (LC-OCD) analysis revealed the abundance of biopolymer in the fouling layer formed at the initial phase. In the second phase, formation of carbonate crystals on the membrane surface was observed but did not affect the flux significantly. In the last phase, the water vapor flux dropped to almost zero due to the deposition of a dense thick layer of sulfate crystals on the membrane surface.

An economic desalination system with a small scale and footprint for remote areas, which have a limited and inadequate water supply, insufficient water treatment and low infrastructure, is strongly demanded in the desalination markets. Here, a direct contact membrane distillation (DCMD) process has the simplest configuration and potentially the highest permeate flux among all of the possible MD processes. This process can also be easily instituted in a multi-stage manner for enhanced compactness, productivity, versatility and cost-effectiveness. In this study, an innovative, multi-stage, DCMD module under countercurrent-flow configuration is first designed and then investigate both theoretically and experimentally to identify its feasibility and operability for desalination application. Model predictions and measured data for mean permeate flux are compared and shown to be in good agreement. The effect of the number of module stages on the mean permeate flux, performance ratio and daily water production of the MDCMD system has been theoretically identified at inlet feed and permeate flow rates of 1.5 l/min and inlet feed and permeate temperatures of 70 °C and 25 °C, respectively. The daily water production of a three-stage DCMD module with a membrane area of 0.01 m at each stage is found to be 21.5 kg.

One of the major challenges in membrane distillation (MD) desalination is scaling, mainly CaSO 4 and CaCO 3. In this study, in order to achieve a better understanding and establish a strategy for controlling scaling, a detailed investigation on the MD scaling was performed by using various analytical methods, especially an in-situ monitoring technique using an optical higher VCF values. In addition, CaCO 3 alone in feed solution did not affect the scaling, however, when CaSO 4 was added to CaCO 3 , the initial MPF decline and VCF met the critical point earlier. In summary, calcium scaling crystal formed at different conditions influenced the filtration dynamics and MD performances.

Developing a high flux and selective membrane is required to make membrane distillation (MD) a more attractive desalination process. Amongst other characteristics membrane hydrophobicity is significantly important to get high vapor transport and low wettability. In this study, a laboratory fabricated carbon nanotubes (CNTs) composite electrospun (E-CNT) membrane was tested and has showed a higher permeate flux compared to poly(vinylidene fluoride-cohexafluoropropylene) (PH) electrospun membrane (E-PH membrane) in a direct contact MD (DCMD) configuration. Only 1% and 2% of CNTs incorporation resulted in an enhanced permeate flux with lower sensitivity to feed salinity while treating a 35 and 70 g/L NaCl solutions. Experimental results and the mechanisms of E-CNT membrane were validated by a proposed new step-modeling approach. The increased vapor transport in E-CNT membranes could not be elucidated by an enhancement of mass transfer only at a given physico-chemical properties. However, the theoretical modeling approach considering the heat and mass transfers simultaneously enabled to explain successfully the enhanced flux in the DCMD process using E-CNT membranes. This indicates that both mass and heat transfers improved by CNTs are attributed to the enhanced vapor transport in the E-CNT membrane.

The low thermal efficiency and low water production are among the major challenges that prevent membrane distillation (MD) process from being commercialized. In an effort to design an efficient multi-stage direct contact MD (DCMD) unit through mathematical simulation, a new phenomenon that we refer to as total water production capacity inversion (WPI) has been detected. It is represented by a decrease in the total water production beyond a number of stages or a certain module length. WPI phenomenon, which was confirmed by using two different mathematical models validated experimentally, was found to take place due to the decrease in water vapor flux across the membrane as well as the increase in heat loss by conduction as the membrane length increases. Therefore, WPI should be considered as a critical MD design-criterion, especially for large scale units. Investigations conducted for a simulated multi-stage DCMD process showed that inlet feed and permeate temperatures difference, feed and permeate flow rates, and feed salinity have different effects on WPI. The number of stages (or module length at constant width) that leads to a maximum water production has been determined for different operating parameters. Decreasing inlet feed and permeate temperatures difference, or inlet feed and permeate flow rates and increasing inlet feed temperature at constant temperature difference or inlet feed salinity cause the WPI to take place at lower number of stages. Even though the feed salinity affects negligibly the 2 mean permeate flux, it was clearly shown that it can affect WPI. The results presented herein unveil a hidden phenomenon that is likely to occur during process scale-up procedures and should be considered by process engineers for a proper choice of system design and operating conditions.

No abstract

This paper presents a theoretical analysis of the monthly average daily and hourly performances of a solarpowered multi-stage direct contact membrane distillation (SMDCMD) system with an energy recovery scheme and dynamic operating system. Mid-latitude meteorological data from Busan, Korea is employed, featuring large climate variation over the course of one year. The number of module stages used by the dynamic operating scheme changes dynamically based on the inlet feed temperature of the successive modules, which results in an improvement of the water production, thermal efficiency, and solar fraction. The simulations of the SMDCMD system are carried out to investigate the spatial and temporal variations in the feed and permeate temperatures and permeate flux. The monthly average daily water production increases from 0.37 m 3 /day to 0.4 m 3 /day and thermal efficiency increases from 31% to 45% when comparing systems both without and with dynamic operation in December. The water production with respect to collector area ranged from 350 m 2 to 550 m 2 and the seawater storage tank volume ranged from 16 m 3 to 28.8 m 3 , and the solar fraction at various desired feed temperatures from 50 °C to 80 °C have been investigated in October and December.