Membrane Distillation Process: Fundamentals, Applications, and Challenges

Abstract

Traditional thermal-based processes such as multistage flash and multi-effect distillation have been used for thousands of years to obtain freshwater from saline water. Recently, with the development of membrane-based technology, membrane distillation (MD) as a thermally driven membrane process has received significant attention. The driving force in MD is the vapor pressure gradient induced by temperature difference through hydrophobic microporous membrane pores. The membrane used for MD should be hydrophobic and microporous. In MD, the mechanism of transport involves simultaneously heat and mass transfers, which moves from the hot feed side to the cold permeate side. The performance of MD is evaluated based on various performance metrics including permeate flux, recovery ratio, thermal efficiency, gained output ratio, and specific thermal energy consumption. It has good ability for various industrial uses due to its moderate applied temperature and pressure, high rejection rate, less membrane fouling tendency and its ability to treat high-saline water. The water production cost still remains high compared to conventional processes. Therefore, MD can be cost-effectively when integrated with solar energy, geothermal energy and waste heat. Nevertheless, MD process requires focused research to improve its efficiency to become more mature and economically competitive at large scale.

1. Introduction

Membrane-based technologies are becoming more attractive for many industrial separation processes, especially in water treatment sector due to their various benefits over thermal-based technologies. The advantages include cost and thermal energy efficiencies, better environmental impacts with less fuel combustion needed to generate thermal energy, and lower pumping and capital costs [1]. Recently, the growing demand for freshwater water has been at high level, explained by the rise of the global water desalination market which was valued at US $15.43 billion in 2017, expected to reach US $27 billion by 2025 [2].

The global water reserve available as freshwater is estimated to be 3%, the remaining 97% is available as seawater. In the coming decades, climate change and population growth are expected to exacerbate the availability of freshwater sources due to shifting rainfall pattern, droughts, temperature pattern, and anthropogenic factors. Non-conventional sources of water, such as desalinated water and recycled/reused wastewater, are considered as the most viable option aimed to alleviate the intensity of water shortage under circular water economy approach. The main membrane-based technological options applied for water desalination and water treatment are microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), membrane bioreactor (MBR), reverse osmosis (RO). However, conventional membrane-based technologies have several limitations including high energy demand, high cost of operating and maintenance, and membrane performance reducing due to fouling phenomena. Conventional thermal-based technologies have been limited by its high need of thermal energy and high cost of operating and maintenance.

Membrane distillation (MD) is an emerging technology that has a consistent margin progression compared to conventional commercialized technologies such as RO, NF, MED, and MSF and has been explored in the field of water desalination and wastewater treatment. MD integrating both thermal distillation and membrane processes. MD is a phase change process, where the driving force is the difference between vapor pressures, induced by temperatures gradient, between two solutions separated by hydrophobic microporous membrane. The main benefit of MD is the ability to operate at lower operating temperature, below its boiling point, compared to conventional thermal-based technologies and at lower hydrostatic pressure compared to conventional pressure-driven membrane technologies. More recently, there has been growing interest in MD technology for its potential to treat high-saline streams with low brine discharge and its aptitude of integration with other membrane-based processes and low-grade heat energies.

Interest in the technology of MD was first proposed by Bodell [3] in 1963, who first explained the process and described the approach to produce potable water. Later, in 1967, Findley [4] published the first documents with interesting results related to MD process. From 1970s to 1990s, MD has gone through long development phase with slow growth. During the last 20 years, there has been a rapid growth with a considerably increase of the number of scientific publications on MD technology, as displayed in Figure 1. Despite the enhanced research interest within the academic community, MD technology is not widely commercialized yet. The main reasons are the lack of suitable designed modules and adequate and the high thermal energy cost.

Figure 1.

Evolution of number of research publications related to MD technology during the period 2000–2022.

In this chapter, fundamental knowledge, membranes and basic configurations, transfer model equations, performance metrics for main applications, and possible integration of the MD system are presented. The understanding of the fundamental mechanisms of MD process is crucial for researchers and engineers in the field of water desalination and water treatment to apply and analyze MD system more effectively. In addition, this chapter provides comprehensive and valuable information on the evaluation of MD performance, which will be a useful guide to studying MD technology.

2. Fundamentals of MD

2.1 Principles of MD

MD is a thermally-driven membrane separation process, only vapor molecules transfer across a microporous hydrophobic membrane [5]. The driving force during MD process is the difference between vapor pressures between two membrane sides generated by a temperature gradient by circulating the hot feed solution (FS) and the cold permeate solution. The vapor generated on the hot FS passes across the membrane and condenses on the cold solution to generate the distillate. The vapor pressure difference increases with the increase of temperature difference between feed and permeate solutions. The main advantage of MD is that considered as low temperature process, it can be operated typically at temperature lower than 80°C [6]. Figure 2 displays the principle of MD process using for the desalination of seawater to produce desalinated freshwater. In the MD process, the hydrophobic nature of the used membrane allows water vapor molecules in the hot feed to pass through the pores of the membrane, while prevents the penetration of the liquid feed water into the membrane pores due to its low surface energy [7]. As a result, a liquid–vapor interfaces are formed at the entrances of the hydrophobic membrane pores. Basically, the pores of the membrane used for MD must be kept dry to prevent the liquid feed water to pass through the hydrophobic membrane.

Figure 2.

The principle of MD.

2.2 Basic MD configurations

In this section, different MD configurations that have been utilized to separate feed solution (FS) using a microporous hydrophobic membrane will be presented. Figure 3 shows the four basic configurations of MD process.

Figure 3.

The four basic configurations of membrane distillation including DCMD: direct contact membrane distillation, VMD: vacuum membrane distillation, AGMD: air gap membrane distillation, and SGMD: sweeping gap membrane distillation.

2.2.1 Direct contact membrane distillation

Direct contact membrane distillation (DCMD) is the simplest and the most studied configuration in MD. In this configuration, the hot FS is in direct contact with one side of the membrane and the cold permeate solution is in direct contact with the other side of the membrane (Figure 3—DCMD). Therefore, solution evaporation takes place at the feed-membrane side and transferred across the membrane porous to condense on the other side in the liquid permeate solution. The hydrophobic characteristic of the membrane prevents the liquid solution to penetrate the membrane porous, only vapor solution can penetrate. The main drawback of the DCMD configuration is its comparatively large conductive heat losses owing to the contact between hot feed side, membrane, and permeate cold side [8].

2.2.2 Vacuum membrane distillation

The schematic diagram of the vacuum membrane distillation (VMD) configuration is presented in Figure 3—VMD. In this configuration, the vapors created in the permeate side are sucked by a vacuum created in the cold side by means of vacuum pump. The condensation takes place outside the module. The applied vacuum pressure should be lower than the saturation pressure created in the hot FS in order to provide sufficient driving force [7]. The advantages of VMD are its high transmembrane water flux and small conductive heat losses. Despite that, VMD presents higher membrane fouling propensity and lower pore-wetting resistance compared to other configurations.

2.2.3 Air gap membrane distillation

In air gap membrane distillation (AGMD), as the schematic diagram shown in Figure 3—AGMD, a stagnant air is introduced between the permeate-membrane side and the condensation surface. The evaporated molecules coming from the hot FS cross both the membrane pores and the air gap to condensate at the cold surface inside the MD module. AGMD shows low heat lost by conduction, which leads to high energy efficiency, but the obtained water flux is generally low due to the additional resistance to mass transfer created by the air gap.

2.2.4 Sweeping gas membrane distillation

The schematic diagram of the VMD configuration is presented in Figure 3—SGMD. In sweeping gas membrane distillation (SGMD), a cold inert gas is applied to sweep the vapor molecules at the permeate-membrane side which is then condensate outside the MD module by means of an external condenser. In this configuration, the sweep gas reduces the mass transfer resistance and consequently enhances the mass transfer coefficient. However, it requires a large sweep gas volume, leading to an increase in condenser capacity. This, in turn, increases additional cost and the system complexity.

2.3 Membrane characteristics

The MD technology is a membrane-based system. The characteristics of the used membranes are of great importance to determine the performance of the process. In MD, the used membranes should be hydrophobic (non-wetting) microporous. Generally, these membranes are made from polyvinylidene fluoride (PVDF), polytetrafluoroethylene (PTFE), or polypropylene (PP) materials. A suitable membrane for MD should have high hydrophobicity, high porosity, uniform pore size distribution, high liquid entry pressure (LEP), low thermal conductivity, high resistance to chemicals, excellent mechanical strength, and low fouling tendency.

Hydrophobic membrane materials should be used to reduce its wettability. The hydrophobicity is evaluated by the measurement of the surface contact angle. The membrane porosity determines the ratio between the volume of the pores and the total volume of the membrane. Higher membrane porosity can guarantee lower conductive heat loss and higher permeate flux. In general, the pore size ranged between 0.1 and 1 μm and the porosity varied between 40% and 90% [8]. The LEP is the minimum hydraulic pressure that must be applied to a liquid before it overcomes the hydrophobicity of the membrane and penetrating the pores [9]. To achieve high MD performance, the value of LEP is suggested to be more than 2.5 bar. A low thermal conductivity is crucial to minimize the heat losses through conduction, hence maintain the driving force for the movement of water vapor across the membrane. Generally, the used materials have thermal conductivity ranged between 0.1 and 0.5 W m⁻¹ K⁻¹. Additionally, high resistance to chemicals (e.g., acids and bases), excellent mechanical strength, and low fouling tendency are crucial characteristics of membranes to ensure MD process performance stability during long-term operation.

2.4 Transport models

2.4.1 Heat transfer

In MD, the mechanism of transport involves simultaneously heat and mass transfers, which occur in the same direction from the hot feed side to the cold permeate side. The heat transfer is carried out in four steps: (1) the heat transferred from the bulk FS to the feed-membrane interface; (2) the heat transferred by conduction through the microporous membrane; (3) the heat associated to the latent heat of vaporization through the membrane pores; and (4) the heat transferred from the permeate-membrane interface to the bulk permeate solution [10, 11].

The heat transferred from the bulk FS to the feed-membrane interface:

$$Q_{\mathrm{f}}=h_{\mathrm{f}}\left(T_{\mathrm{f}}-T_{\mathrm{fm}}\right)$$E1

The heat transferred by conduction through the membrane associated with the heat associated to the latent heat of vaporization through the membrane pores:

$$Q_{\mathrm{m}}=Q_{\mathrm{v}}+Q_{\mathrm{c}}=J∆H_{\mathrm{v}}+h_{\mathrm{m}}\left(T_{\mathrm{fm}}-T_{\mathrm{pm}}\right)$$E2

The heat transferred from the permeate-membrane interface to the bulk permeate solution:

$$Q_{\mathrm{p}}=h_{\mathrm{p}}\left(T_{\mathrm{p}}-T_{\mathrm{pm}}\right)$$E3

Where h_f, h_p, and h_m are the heat transfer coefficient in feed side, permeate side, and membrane respectively. T_f and T_p are the bulk feed and bulk permeate temperatures. T_fm and T_pm are the temperature at the feed-membrane and permeate-membrane interfaces, respectively. Q_v and Q_c are the heat energy of vaporization and heat of condensation, respectively. J is the mass transfer of vapor through the membrane porous. ΔH_v is the latent heat of vaporization.

The overall heat transfer flux through the whole MD system is given by the following equation:

$$Q=Q_{\mathrm{f}}=Q_{\mathrm{p}}=Q_{\mathrm{m}}$$E4

2.4.2 Mass transfer

The mass transport in MD involves three major stages that occur in series: (1) vapor generation takes place at the feed-membrane interface; (2) vapor transport across the membrane pores; and (3) vapor condensation at the permeate-membrane interface. The mass transfer of vapor through the membrane porous is generally defined according to Darcy’s law, expressed as follows [12]:

$$J=B_{\mathrm{m}}\left(P_{\mathrm{m},\mathrm{f}}-P_{\mathrm{m},\mathrm{p}}\right)$$E5

Where B_m is mass transfer coefficient, P_m,f is the partial vapor pressure at the feed-membrane interface, and P_m,p is the partial vapor pressure at the permeate-membrane interface.

According to the dusty gas model (DGM), the prediction of mass flux through a porous membrane can be controlled by three basic mass transfer mechanisms, including Knudsen diffusion, viscous or Poiseuille flow, molecular or ordinary diffusion [13]. To determine the dominant mass transfer mechanism of flow through the membrane pores, it is necessary to calculate the Knudsen number (K_n), from the following equation:

$$K_{\mathrm{n}}=\frac{\lambda }{d_{\mathrm{p}}}$$E6

Where d_p is the pore diameter and λ is the mean free path, which can be calculated from the following equation:

$$\lambda =\frac{k_{\mathrm{B}}{T}_{\mathrm{m}}}{\pi \ast {\left(\frac{{\sigma }_{\mathrm{w}}-{\sigma }_{\mathrm{a}}}{2}\right)}^{2}P\sqrt{1+\left(\frac{M_{\mathrm{w}}}{M_{\mathrm{a}}}\right)}}$$E7

Where k_B is the Boltzmann constant. T_m is the average temperature of the membrane. σ_w and σ_a are the collision diameters of water and air; respectively. P is the total pressure inside the pore. M_w and M_a are the molecular weights of the water and the air, respectively.

If the Knudsen number is less than 0.01, the mass transfer mechanism is dominated by the molecular diffusion. If the Knudsen number is ranged between 0.01 and 1, the mass transfer mechanism is dominated by the combined Knudsen-molecular diffusion. If the Knudsen number is larger than 1, the mass transfer mechanism is dominated by the Knudsen diffusion [14].

Therefore, the flux of vapor through the membrane porous can be calculated using the following equations for molecular, Knudsen-molecular, and Knudsen diffusion, respectively:

For molecular diffusion:

$$J=\frac{\epsilon \ast D_{\mathrm{w}}\ast M_w}{\tau \ast \delta \ast P_{\mathrm{a}}\ast R\ast T}\left(P_{\mathrm{m},\mathrm{f}}-P_{\mathrm{m},\mathrm{p}}\right)$$E8

For Knudsen-molecular diffusion:

$$J={\left[\frac{3\ast \tau \ast \delta }{\epsilon \ast d_{\mathrm{p}}}\sqrt{\frac{\pi \ast R\ast T}{8\ast M_{\mathrm{w}}}}+\frac{\tau \ast \delta \ast P_{\mathrm{a}}\ast R\ast T}{\mathrm{\epsilon }\ast P{D}_{\mathrm{w}}\ast M_{\mathrm{w}}}\right]}^{-1}\left(P_{\mathrm{m},\mathrm{f}}-P_{\mathrm{m},\mathrm{p}}\right)$$E9

For Knudsen diffusion:

$$J=\frac{\epsilon \ast d_{\mathrm{p}}}{3\ast \tau \ast \delta }\sqrt{\frac{8\ast M_{\mathrm{w}}}{\mathrm{\pi }\ast R\ast T}}\left(P_{\mathrm{m},\mathrm{f}}-P_{\mathrm{m},\mathrm{p}}\right)$$E10

Where ε is the membrane porosity, τ is the membrane tortuosity, δ is the membrane thickness, D_w is the diffusion coefficient of water, P_a is pressure of air, P is the total pressure inside the pore, R is the gas constant, and T is the temperature.

3. Membrane distillation metrics

Some performance indicators have been used to evaluate the efficiency of MD process such as permeate flux (J_p), recovery rate (Y), thermal efficiency (η), gained output ratio (GOR), and specific thermal energy consumption (STEC).

3.1 Permeate flux

Permeate flux is the main parameter used as performance indicator for MD process, it presents the productivity of the system. The permeate flux can be defined as the water flow rate passing through a specified membrane area at a certain time. It is calculated using the following equation:

$$J_p=\frac{m_p}{S\ast \mathit{\Delta t}}$$E11

Where m_p is the produced water weight, S is the effective membrane surface, and ∆T is the time interval.

3.2 Recovery ratio

The recovery ratio is a significant performance criterion which is used to determine the design size and the economic aspect of MDs system. It is defined as the percentage of the produced permeate flow rate over the feed flow rate.

$$Y=\frac{m_{\mathrm{p}}}{m_{\mathrm{f}}}\ast 100$$E12

Where m_f is the feed water weight.

3.3 Thermal efficiency

In MD, the heat transfer is divided into two steps including heat transfer through membrane matrix by conduction and heat transfer by movement of vapor across the membrane induced by latent heat of vaporization. The heat transfer through membrane matrix by conduction is considered as heat loss, which must be minimized to improve the thermal efficiency of the membrane. Thermal efficiency is defined as the ratio of the useful energy delivered (latent heat of vaporization) to the total heat (latent and conduction) and is calculated using equation as follows:

$$\eta =\frac{J_{\mathrm{p}}∆H_{\mathrm{v}}}{J_{\mathrm{p}}∆H_{\mathrm{v}}+\left(\frac{k_{\mathrm{m}}}{\delta }\right)\left(T_{\mathrm{fm}}-T_{\mathrm{pm}}\right)}\ast 100$$E13

Where ΔH_v is the latent heat of vaporization, T_fm and T_pm are the temperature at the feed-membrane and permeate-membrane interfaces, respectively, k_m is the thermal conductivity of the membrane, δ is the membrane thickness.

3.4 Gained output ratio

The GOR is used to assess the energy efficiency of MD system. GOR can be defined as the ratio of the thermal energy required to vaporize the mass of water produced to the total thermal energy provided to the system [15]. GOR is calculated using the following equation:

$$\mathrm{GOR}=\frac{F_{\mathrm{d}}S{\rho }_{\mathrm{d}}{\mathit{\Delta H}}_{\mathrm{v}}}{F_{\mathrm{f}}{\rho }_{\mathrm{f}}{C}_{\mathrm{p},\mathrm{f}}\left(T_{\mathrm{f},\text{in}}-T_{\mathrm{p},\text{out}}\right)}$$E14

Where F_d and F_f are the distillate water and feed water flow rates, respectively. ρ_d and ρ_f are the density of distillate and feed waters, respectively. T_f,in and T_p,out are the feed inlet and the permeate outlet temperatures. ΔH_v is the latent heat of vaporization. C_p,f is the specific heat of feed water. S is the effective membrane surface.

3.5 Specific thermal energy consumption

The STEC is another thermal performance indicator for MD process. STEC represents the amount of external thermal energy consumed by MD system to produce a unit volume of distilled water and is calculated using the following equation:

$$\mathit{\text{STEC}}=\frac{F_{\mathrm{f}}\ast {\rho }_{\mathrm{f}}\ast C_{\mathrm{p},\mathrm{f}}\ast \left(T_{\mathrm{f},\text{in}}-T_{\mathrm{p},\text{out}}\right)}{3.6\ast {10}^6\ast F_{\mathrm{d}}}$$E15

4. Potential applications

The main advantages of MD are its ability to achieve high rejection rate (theoretically 100%), and it operates at atmospheric pressure and relatively low temperatures. These supremacies of MD compared to conventional membrane-based technologies, render it a promising technology which gained significant attention for various potential applications.

MD can be applied for the desalination of saline water including seawater and brackish water. El Mokhtar et al. [16] investigated the feasibility of AGMD for the desalination of seawater. At feed temperature of 77°C, AGMD process produced freshwater with high salt rejection, more than 99%, and permeate flux of 9.06 kg m⁻² h⁻¹. The comparison between experimental and predicted models shown high accuracy. Another study conducted by Usman et al. [17] provided a systematic evaluation of the economics involved in desalination of brackish water using MD process to produce potable water. The combination of solar-thermal sources with waste heat to power MD system greatly lowered the water production cost. MD system relying on waste heat and solar thermal can lower water production cost from $6.80/m³ to a mere $1.6/m³. Additionally, MD is less sensitive to high salinity, it can be used for the desalination of hypersaline solutions such as brine and concentrated wastewater toward near-zero liquid discharge.

Furthermore, high removal efficiency of small molecule contaminants and heavy metal ions is achieved by MD. Lou et al. [18] investigated the efficiency of MD process integrated with crystallization technology for the treatment of concentrated heavy metal wastewater, including zinc and nickel. Results show that MD process integrated with crystallization proves to be a promising technology for treating highly concentrated heavy metal solutions and the used membrane has excellent resistance to fouling following the treatment of highly concentrated solutions.

MD also has the potential to minimize the amount of wastewater discharged from various wastewater sources, such as municipal wastewater, textile wastewater, pharmaceutic industry wastewater, liquid nuclear wastewater, oily wastewaters, etc. Produced water, is a by-product wastewater effluent with high salinity, contains suspended particles, dispersed oils, and chemicals. MD can be applied for the treatment of highly saline produced water. MD also has the potential to recover valuable components. It can be utilized for the concentration of fruit juices, herbal extracts, sugars, alcohols, mineral acids, etc. Criscuoli and Drioli [19] evaluated the performance of VMD process to concentrate date juices at low temperature. After previous treatment steps (enzymatic treatment, clarification, decolorization, and deionization), the date juices can be processed by VMD process with low membrane fouling tendency.

5. MD-based hybrid processes

5.1 Integration with membrane-based processes

5.1.1 RO-MD

Pressure-driven RO is highly efficient desalination technique for production of freshwater. However, it is limited by its low capability to treat high concentrated streams due to physical limitations caused by high osmotic pressure values. The RO water recovery for the desalination of seawater is limited to 60% when using a double stage and applying high pressure. In this case, MD is integrated to treat the RO brine to overcome the limitation of water recovery. Hence, RO-MD integrated system can increase the overall water recovery and therefore reduce the brine discharge to achieve minimal liquid discharge approach. However, the main drawback of coupling MD with RO is the high MD membrane scaling propensity induced by the presence of high chemicals content in RO brine (salts, antiscalants, chemical cleaning, etc.)

5.1.2 FO-MD

Forward osmosis (FO) is an osmotically-driven separation process that utilizes the osmotic pressure difference between a FS and a draw solution (DS) to transport water molecules across a semi-permeable membrane [20]. It is considered as the lower required energy consumption among the desalination processes. The main drawback of FO is that it is limited by the dilution of DS, thus a recovery process must be added [5]. For this purpose, MD can be integrated with FO process to provide a good alternative for the regeneration and the reuse of DS as well as the generation of high-water quality. In FO-MD hybrid process, the FS is concentrated by the FO system, and the diluted DS is concentrated by the MD process to produce high osmotic pressure solution. The FO-MD integrated system can be applied for water desalination and wastewater treatment with high-water recovery and shown great potential in achieving zero liquid discharge approach.

5.1.3 MBR-MD

MBRs combine conventional bioreactor with pressure-driven UF or MF membranes under aerobic and anaerobic conditions. MBR is mainly used to separate suspended solids from domestic and industrial wastewater due to its simplicity, reliability, and cost-effectiveness. However, the used membranes are inefficient in retaining inorganic salts and trace organic contaminants, thus the process cannot attain water reclamation and reuse [21]. Therefore, MD can be integrated with MBR to form the hybrid membrane distillation bioreactors (MDBR) for efficient wastewater treatment and recovery. MDBR integrated system can be potentially used for water, nutrients, bioenergy, and recovery of value-added products from domestic and industrial wastewaters.

5.2 Integration with renewable energy

5.2.1 Solar energy

The lower operating temperatures needed for MD technology compared to conventional thermal processes make it possible to be coupled with alternative low-grade heat sources such as solar energy, geothermal energy, and waste heat. Using alternative energy sources can drastically lower the cost and energy consumption of the process, as well as reduce its environmental footprint.

Solar energy is widely used as renewable energy sources for thermal energy generation in MD. It used energy collected from sunlight to heat the FS for MD process. Collecting heat from solar energy can be generated by two methods: direct collecting and indirect collecting [22]. The direct collecting method consists of converting concentrated solar radiation into thermal energy through a moving fluid, as a kind of heat exchanger. The indirect collecting method consists of converting solar energy into electricity through photovoltaic cells. The generated electricity is then used directly or stored in batteries for heating the FS or operate various components in MD system such as pumps, compressors, etc. It can be reported that using solar energy-powered MD system instead of electricity-powered system, under the same conditions, can reduce the specific energy consumption by 30% and increase the GOR by 17% [23].

5.2.2 Geothermal energy

Unlike solar energy, geothermal energy is found in deep underground and does not depend on weather and climatic conditions. Geothermal energy sources are expected to reduce the energy cost and lower the overall water production cost in MD system. The main advantage of using geothermal energy in MD is its stable heat source that is readily accessible, enabling the direct use for heating feed water without the need for an energy converter. The integration of low-grade heat sources of geothermal energy with MD is potentially applied for the desalination of geothermal brackish water and wastewater treatment. The scaling up potential of MD is related to its capability to use the steady heat source of geothermal energy which is better than using solar energy [10]. Despite the abundance of geothermal energy sources in many regions in the world, the integration of this kind of renewable energy is limited due to the limitation of adequate geographical sites.

5.2.3 Waste heat

Waste heat is a freely available low-grade energy source released at a temperature lower than 100°C from various industries such as power stations, diesel engines, nuclear reactors, and industrial plants, etc. As waste heat is a freely abundant energy source, it can be integrated with MD which improves the energy efficiency of the system with no additional carbon footprint. Thus, coupling MD with waste heat is a promising solution and becomes environmentally and economically competitive with conventional membrane-based technologies. The low-grade waste heat is integrated with MD used for the desalination of seawater and brackish water. It can be reported that MD powered by waste heat source reduces the energy cost to 0.31 USD/m³, which is lower than those using conventional RO technology with 0.45 USD/m³ [24].

6. Challenges in MD process

6.1 Polarization phenomena

6.1.1 Temperature polarization

In MD, the low thermal efficiency generated by temperature polarization is one of the key obstacles hampering its applications. The vaporization occurs at the membrane hot surface and the condensation occurs at the membrane cold surface, leading to the creation of a thermal boundary layer between two membrane sides. This effect means that a large amount of heat supplied to the process is transferred to the cold side by conduction instead of evaporating feed water. This phenomenon is called temperature polarization, estimated using the temperature polarization coefficient (TPC), which is defined as follows:

$$\mathrm{TPC}=\frac{T_{\mathrm{fm}}-T_{\mathrm{pm}}}{T_{\mathrm{f}}-T_{\mathrm{p}}}$$E16

Where T_fm is the temperature on the feed-membrane side, T_pm is the temperature on the permeate-membrane side, T_f is the temperature of the bulk feed, and T_f is the temperature of the bulk permeate.

6.1.2 Concentration polarization

Like temperature polarization, concentration polarization takes place in MD process, but it has less effect on permeate flux compared to temperature polarization. MD process provides the theoretically 100% rejection of all non-volatiles, indicating that the concentration polarization only occurs in the feed side. An increase in concentration polarization led to an additional mass transfer resistance that will decline the permeate flux. The estimation of the concentration polarization is calculated by concentration polarization coefficient (CPC), which can be expressed as follows [25]:

$$\mathrm{CPC}=\frac{C_{\mathrm{fm}}}{C_{\mathrm{f}}}$$E17

Where C_f,m is the solute concentration on the feed-membrane side and C_f, is the solute concentration in the bulk feed.

6.2 Membrane fouling

Membrane fouling is the accumulation of substances on the membrane surface or inside the membrane pores, which causes a deterioration of the overall performance of MD in terms of permeate flux and solute rejection. Generally, membrane fouling can be classified into three categories according to the source of feed water, including inorganic fouling or scaling, organic fouling, and biological fouling. Inorganic fouling is caused by the accumulation of salt precipitates such as calcium sulfate, calcium carbonate, silicate, and sodium chloride, etc. Organic fouling is caused by colloidal organic matters such as humic substances, extracellular polymeric substances, and proteins. Biological fouling (biofouling) is caused by the build-up of microorganisms on the membrane surface [26]. The main parameters that affect membrane fouling are the depositing materials characteristics (concentration, solubility, diffusivity, charge, etc.), and membrane structure (hydrophobicity, roughness, pore size, charge, etc.).

6.3 Membrane wetting

Membrane wetting is one of the main challenges faced in the development and commercialization of MD technology. Membrane wetting is a phenomenon in which the liquid FS freely permeates through the pores of the hydrophobic membrane, which causes a deterioration of membrane performance in terms of solute rejection. Membrane wetting is caused by the presence of surface-active materials, such as oils or surfactants which adsorb onto the hydrophobic MD membrane and make it progressively more hydrophilic [27]. In MD, the membrane-wetting phenomenon is detected simply by a simple measurement of distillate electrical conductivity.

7. Conclusion and future directions

MD is an emerging thermally-driven membrane-based system that has become one of the most attractive processes in sustainable techniques to help reduce the global water-energy stress in separation and water treatment applications. The integration of MD with conventional membrane-based processes such as RO, FO, and MBR, can significantly enhance its performance. Moreover, the integration of renewable and low-cost energy sources such as solar energy, geothermal energy, and waste heat can make MD technology cost- and energy-effective with more sustainable way. MD has the advantage to have higher solute rejection rate than pressure-driven technologies and lower thermal energy consumption than conventional distillation technologies. MD also can treat high-saline waters with high recoveries generated near-zero liquid discharge approach applied for water desalination and wastewater purification. However, MD is still facing many challenges that hinder its wide industrial application, due to the relatively low water flux and high energy consumption conventional RO technology. Therefore, future directions require focused research to improve the efficiency of MD to become more competitive at large scale, among them:

1. Design and fabrication of appropriate membrane and modules for MD application. The purpose of this approach aimed to reduce the effect of temperature and concentration polarization in order to enhance the efficiency of the MD process. Membranes with high thermal conductivity, large pore size, high porosity, low tortuosity, small thickness are required to achieve better performance. Moreover, lack of suitable designed modules limits the deployment of MD technology at industrial scale. Exploring new module configurations suitable for the specificity of MD flows is also interesting.

2. Reduce the energy consumption for large-scale translation of MD technology to achieve sustainable water production. Renewable energies or waste heat-powered MD are well studied in lab-scale, but extension and optimization for large-scale remain essential. Low-grade heat coming from solar energy, geothermal energy, and waste heat is suitable for MD system. However, the location where the MD process would be implemented is a primordial factor for successful operation.

3. Reduce membrane fouling tendency and pore-clogging in order to extend the membrane lifespan. Membrane fouling is a dynamic process that is affected by various factors including feed water quality, membrane properties, and operating conditions. Develop an in-depth understanding of fouling mechanism to make a prevention strategy to mitigate membrane fouling and improve the performance of used membrane.