Desalination of brackish groundwater by solar-powered membrane distillation

Freshwater scarcity has become one of the major challenges humans are faced with in this century. Natural freshwater is becoming scarce while water demand is continuously rising with the rapid population and economic growth in many regions of the world. Desalination of brackish water and seawater is more and more considered as an alternative way of freshwater supply. The utilization of solar energy to drive desalination processes is a potential sustainable solution to the world’s water scarcity issue. This study comprehensively reviewed recent development and applications of seven major solar desalination technologies, including solar still, solar humidification-dehumidification process, solar powered multi-stage flash (MSF), multi-effect distillation (MED), membrane distillation (MD), reverse osmosis (RO) and electrodialysis (ED). Potential integration of solar technologies and desalination processes are summarized. By collecting and analyzing performance data from recent studies, the status of productivity, energy production costs of different technologies are critically reviewed. The real-world applicability, as well as the technical and economic feasibility of these technologies, is also evaluated.

Brackish groundwater treatment by solar-powered membrane distillation is a potential sustainable alternative for drinking water supply in many arid, semi-arid regions that lack freshwater resources. In this research, a three-loop solar-powered vacuum membrane distillation (SVMD) has been developed, which was designed to be operated autonomously with abundant solar radiation. Evacuated tube solar collectors and PV panels are utilized to supply thermal and electrical energy, respectively. A commercial polypropylene capillary membrane module with 0.1 m2 effective surface area is adopted in the VMD process. In order to improve the system efficiency, the SVMD setup is continuously optimized during the implementation of the experimental studies. Major modifications include the installation of a new shell-and-tube condenser, higher capacity pumps in the membrane feeding and cooling loops, and the optimization of the permeate side pipelines. The effect of these modifications on the SVMD performance is investigated and discussed.

Brackish groundwater (BGW) has a TDS range of 600~30,000 mg/L. Besides the high salinity, BGW often contains some naturally occurring inorganic contaminants, such as fluoride, nitrate, iron, and manganese. The removal of these contaminants from brackish groundwater by the SVMD system was investigated under real weather conditions. High quality permeate water with TDS of less than 12 mg/L was produced and more than 99.7% salt rejection rate was achieved in the tests. Fluoride was not detected in any of the permeate water samples while nitrate, iron and manganese concentrations were all below the Australian Drinking Water Guideline. Modifications of the SVMD system were done between the three experimental phases of this study. Significant improvement of permeate flux and overall thermal efficiency was obtained after the modification.

Membrane scaling and wetting is one of the major problems in the application of membrane distillation technology. The scaling and wetting of the membrane during brackish groundwater treatment were investigated in this study using the SVMD system with the commercial PP capillary membrane module as well as using a DCMD system with commercial flat sheet PTFE membrane. The SVMD experiment was conducted under real weather conditions for 10 days while the DCMD experiment was operated for 188 h intermittently. Significant permeate flux decline and severe membrane wetting were observed in both cases. A dramatic decline of salt rejection rate, i.e. from almost 100% to 60%, was observed on the last day of SVMD operation. The exceedance of LEP during the SVMD operation was supposed to be the major reason for the membrane wetting. Acid cleaning and membrane drying were able to effectively recover the permeate flux and temporarily mitigate membrane wetting. However, the hydrophobicity of the membrane was not fully recovered. Membrane wetting occurred gradually in the DCMD experiment, resulting in an increase of daily average permeate conductivity, i.e. from less than 3.0 μS/cm to 160 μS/cm. A 57% reduction of water contact angle was observed for the fouled PTFE flat sheet membrane after the DCMD experiment. The feed side of the PP capillary membrane in the SVMD system also showed 19% lower contact angle than the relatively unaffected permeate side PP membrane after the cleaning and drying procedure. SEM/EDS results revealed that the fouled PTFE membrane surface was almost covered by CaCO3 salt crystals. While in the PP capillary membrane examined after acid cleaning, only a small amount of deposit was detected on the feed side membrane surface. Cu, Zn and Fe oxides were found to be the major components of the remaining foulants.

A few scaling mitigation approaches were examined for brackish groundwater (BGW) treatment with direct contact membrane distillation, including regular distilled water flushing, pre-acidification and degassing, and addition of a polyacrylic acid based antiscalant X1030A into the feed water. During the 200 h DCMD operation, all three approaches have reduced the permeate flux decline to some degree. Among them, the antiscalant addition group achieved the most stable permeate flux performance. However, membrane wetting occurred during the operation. The daily average permeate water conductivity gradually increased from less than 4.0 μS/cm to 52.4 μS/cm. Membrane wetting did not occur in DCMD experiments without or with the other two scaling mitigation approaches. Less permeate flux reduction at low conductivity levels, less deposits in the SEM images and higher water contact angles on the fouled membrane were observed in the pre-acidification + degassing group than the regular distilled water flushing group.

Chemical cleaning of the fouled and wetted PTFE membrane after 100 h DCMD BGW treatment was conducted using five different cleaning agents (i.e. deionized water, 3 wt.% HCl solution, 5 wt.% citric acid, 0.2 wt.% EDTA solution and 0.1 wt.% oxalic acid + 0.8 wt.% citric acid). 3 wt.% HCl showed the best performance in terms of scale removal and membrane feed side contact angle recovery, followed by 0.2 wt.% EDTA solution and 5 wt.% citric acid. However, long-term or frequent 3 wt.% HCl cleaning may potentially have a negative impact on the hydrophobicity of the membrane.