Solar energy assisted direct contact membrane distillation (DCMD) process for seawater desalination

Highlights

• Solar assisted DCMD system for seawater desalination.

• Improved two dimensional dynamic model for unsteady state conditions and membrane fouling.

• Model predicts the permeation flux with time for different operating conditions.

• Solar-MD system proved suitable for long term operation of approximately around a half year.

Abstract

Development of a solar assisted direct contact membrane distillation (DCMD) system for seawater desalination and an improved mathematical model to predict the permeate flux for unsteady state conditions were investigated. Different types of commercially available polytetrafluoroethylene (PTFE) membranes were used in a solar-DCMD system for seawater desalination. Membrane properties, such as the liquid entry pressure (LEP), pore diameter, effective porosity and pore size distribution, were characterized for each membrane. A two dimensional (2D) flat-plate dynamic model with heat and mass transfer mechanisms was used to predict the permeate flux under different operating conditions. Good agreement between the numerical simulation and experimental results were found. Long-term fouling phenomenon in the DCMD system was experimentally and theoretically examined. The experimental heat energy consumption ranged from 896 kW h/m³ to 1433 kW h/m³, and the gained output ratio (GOR) ranged from 0.44 to 0.70. The solar-DCMD system was run continually for more than 150 days for seawater desalination in Korea. During day time, more than 77.3% of the heating energy was supplied by solar energy. In particular, in the month of September, 95.3% of the heating energy was supplied by solar energy.

Introduction

Membrane distillation (MD) is an advanced membrane based separation technology, which producing high purity water from salty water such as seawater, which is key potential application for MD. Currently, energy consumption is one of the main challenges for MD desalination, which was estimated to be more than 450 kW h/m³ [1]. In comparison, the energy consumption for reverse osmosis (RO), multiple effect distillation (MED) and multi-stage flash (MSF) are approximately 7 kW h/m³, 40 kW h/m³ and 40 kW h/m³, respectively, leading to a relatively high cost for water production for membrane distillation [2], [3], [4], [5]. On the other hand, MD only requires a moderate temperature to generate the thermal driving force across the membrane, which makes it viable to utilize waste heat or renewable solar energy to reduce the water production cost [1]. Furthermore, the salt concentration has relatively little effect on the mass flux in the MD process compared to the RO process, indicating that MD can deal effectively with high concentration brine [6]. This may be economically competitive when low-grade waste heat or renewable energy resources, such as solar energy, are available [7], [8].

In this research, the solar energy was collected by a solar water heater to supplying the heating energy for DCMD process. Insolation is a measure of the solar radiation energy received on a given surface area during a given time. To date, there are no generally agreed-on norms or standards to assess the performance of solar desalination systems. The parameter normally used to rate the performance of thermal desalination processes is the gained output ratio (GOR). The GOR is defined as the energy ratio of the latent heat of evaporation of the product water to the input thermal energy [9],

$$\text{GOR}=\frac{\mathrm{\Delta }{h}_{\mathit{evap}}{m}_{\mathit{product}}}{Q_{\mathit{input}}}$$

where h_evap is the heat for water saturated vapor, which was obtained as follows [10]:

$$h_{\mathit{evap}}=1850.7+2.8273T-1.6\times {10}^3{T}^2$$

and

$$Q_{\mathit{input}}=m_h{C}_{\mathit{pF}}(T_{\mathit{hi}}-T_{\mathit{ho}})$$

where m_h is the hot stream flow rate, C_pF is the heat capacity of feed solution, T_hi is the feed inlet temperature and T_ho is the feed outlet temperature. Higher GOR systems, therefore, represent more efficient thermal energy systems and consume less energy. Hence, they have lower energy requirements and costs, and for the case of solar energy systems, the solar collection area is reduced representing reduced capital investment. The GOR of the membrane modules reported in literature has ranged from 0.3 to 6 (106–2100 kW h/m³) [11], [12], [13].

Another objective of this study is to build a more reliable and comprehensive mathematical model for MD process. Most MD models are developed assuming the process as one-dimensional using empirical heat and mass transfer equations [14], [15], [16], [17], [18], [19]. Two-dimensional (2D) theoretical models have been considered only in recent years, in which the feed and permeate temperatures along the longitudinal (x-axis) and transversal (y-axis) directions have been considered, and local permeate fluxes as well as total flux have been determined [20], [21], [22], [23]. The Navier–Strokes equations in 2D domains have been employed to provide more reliable information on flow fields. For example, Charfi et al. used the continuity, momentum, energy, and mass transport equations to examine the heat and mass transfer in sweeping gas membrane distillation process [21]. Our previous published paper used numerous submodels, such as the dusty-gas model, momentum, energy, and mass transport equations to study the mass and heat transfer in DCMD systems [24]. In this investigation, a 2D mathematical model was extended to consider simulation for unsteady state conditions. There has been very little reported for unsteady state modeling of DCMD processes [25]. In order to predict the flux for long term operation, the following were considered: (1) a time factor inside the heat transfer equation to account for changes in heat transfer conditions arising from fouling; (2) similarly a time factor was also used for flux prediction to account for fouling which has only previously been considered by a few people [26], [27], [28]; and (3) different types of commonly used mass transfer models were compared with the experimental results.

Previous studies examined the following: the effect of the operating parameters and membrane types for an AGMD system [29], [30], a DCMD system in long term fouling trials [31], and the effect of the module dimensions on the membrane flux [24]. In the present study, a solar assisted DCMD system was constructed for seawater desalination. Three different types of polytetrafluoroethylene (PTFE) membranes from different companies were studied for comparison. The membrane liquid entry pressure (LEP), pore diameter, effective porosity and pore size distribution were characterized and used in the analysis. The DCMD process energy consumption and GOR were studied and calculated. Seawater was run continually through the solar energy assisted DCMD system for more than 150 days to allow an examination of the fouling phenomena and membrane fouling coefficient.