Comparative Energetics of Various Membrane Distillation Configurations and Guidelines for Design and Operation

Abstract

This paper presents a comparative performance study of single-stage desalination processes with major configurations of membrane distillation (MD) modules. MD modules covered in this study are (a) direct contact MD (DCMD), (b) vacuum MD (VMD), (c) sweeping gas MD (SGMD), and (d) air gap MD (AGMD). MD-based desalination processes are simulated with rigorous theoretical MD models supported by molecular thermodynamic property models for the accurate calculation of performance metrics. The performance metrics considered in MD systems are permeate flux and energy efficiency, i.e., gained output ratio (GOR). A general criterion is established to determine the critical length of these four MDs (at fixed width) for the feasible operation of desalination in a wide range of feed salinities. The length of DCMD and VMD is restricted by the feed salinity and permeate flux, respectively, while relatively large AGMD and SGMD are allowed. The sensitivity of GOR flux with respect to permeate conditions is investigated for different MD configurations. AGMD outperforms other configurations in terms of energy efficiency, while VMD reveals the highest permeate production. With larger MD modules, utilization of thermal energy supplied by the hot feed for evaporation is in the order of VMD > AGMD > SGMD > DCMD. Simulation results highlight that energy efficiency of the overall desalination process relies on the efficient recovery of spent for evaporation, suggesting potential improvement in energy efficiency for VMD-based desalination.

Keywords: e-NRTL; energetics; gained output ratio (GOR); membrane distillation (MD); modeling.

Figure 5

Flowchart of a base case DCMD-based desalination process.

Figure A1

DCMD model validation (experimental data can be retrieved from [17]).

Figure A2

VMD model validation (experimental data can be retrieved from [16]).

Figure A3

AGMD model validation (experimental data can be retrieved from [15]). (a) Impact of feed temperature and (b) air gap thickness on the membrane flux.

Figure A4

Flowchart of a base case VMD-based desalination process.

Figure A5

Flowchart of a base case SGMD-based desalination process.

Figure A6

Flowchart of a base case AGMD-based desalination process.

Figure 1

Control volumes of feed and permeate sides for different MD modules. Mass, momentum, and energy balances are made over the control volume of thickness $\mathsf{\Delta }z$. $J_w$ and $q_f$ represent mass and heat flux through the membrane, respectively; $T_{f,b}$ and $T_{p,b}$ feed and permeate bulk temperature, respectively; $T_{f,m}$ and $T_{p,m}$ membrane surface temperature on the feed and permeate sides, respectively; $T_p$ temperature on the permeate side; $T_{m,g}$, $T_{g,fl}$, $T_{fl,pl}$, $T_{pl,c}$, and $T_{c,b}$ temperature at the membrane-air gap interface, the air gap-condensate film interface, the condensate film-cooling plate interface, and the cooling plate-coolant interface, respectively; $C_{f,b}$ and $C_{f,m}$ concentration in the feed bulk solution and on the feed side of the membrane, respectively; $v_{p,in}$ and $v_{p,out}$ permeate velocity at the inlet and outlet, respectively; $v_{a,in}$ and $v_{a,out}$ air velocity at the inlet and outlet, respectively; $v_{c,in}$ and $v_{c,out}$ coolant velocity at the inlet and outlet, respectively; and $P_p$ vacuum on the permeate side.

Figure 2

Saturated temperatures of water equivalent to vapor pressure depression, moisture content of air, and vacuum pressure. The values of a and b are 0.988 and 0.0492, respectively, obtained from the exponential curve fitting of vapor pressure vs. temperature from the Antoine equation.

Figure 3

Flowchart for performance calculation of MD-based desalination.

Figure 4

Variation in $\mathsf{\Delta }{T}_m$ (solid lines—left y-axis) and $\mathsf{\Delta }{T}_w$ (dashed lines—right y-axis) along the length at different feed salinities for (a) DCMD, (b) VMD, (c) SGMD, and (d) AGMD.

Figure 6

GOR–flux performance for permeate condition in DCMD with varying distillate flow rate of 1200–2000 kg/h and temperature of 20–40 °C. Note that GOR and flux are calculated from the DCMD-based desalination process at each condition.

Figure 7

GOR–flux performance for permeate pressure in VMD with varying permeate pressure of 3.5–8.5 kPa and module length of 1–3 m. Note that GOR and flux are calculated from the VMD-based desalination process at each condition.

Figure 8

GOR–flux performance for permeate conditions in SGMD with varying (a) airflow rate of 400–1300 kg/h and (b) RH of 0–90%. Note that GOR and flux are calculated from the SGMD-based desalination process at each condition.

Figure 9

GOR–flux performance for permeate conditions in AGMD with varying coolant temperature of 20–50 °C for (a) air gap thickness of 1–4 mm and (b) coolant flow rate of 1250–2000 kg/h. Note that GOR and flux are calculated from the AGMD-based desalination process in each condition.

Figure 10

Vapor permeabilities in different mass transport regions of the AGMD process. C_KN, C_Mol, C_mem, C_gap, and C_ov denote the vapor permeability corresponding to Knudsen diffusion, molecular diffusion, diffusion inside the porous membrane, mass transfer through the air gap, and overall mass transfer through the membrane and air gap.

Figure 11

Flux, vapor permeability, and vapor pressure difference profile with varying gap thickness.

Figure 12

Thermal efficiency ($\eta$) of different MD configurations.

Figure 13

Partial condensation of vapor in VMD and SGMD-based desalination process.

Figure 14

Flux vs. STEC in MD-based desalination.

Figure 15

Qualitative summary of various MD configurations versus various key efficiency, operating, and economic parameters.