Impacts of silica on the sustainable productivity of reverse osmosis membranes treating low-salinity brackish groundwater

Abstract

Brackish water desalination is increasingly being considered as an alternative drinking water treatment processes. The impacts of high silica (63 mg/L) on the performance of brackish water reverse osmosis (BWRO) while treating a brackish groundwater (1315 mg/L of TDS) were investigated. This study demonstrated the importance of pilot-scale testing in understanding the long-term impacts of silica fouling in order to maintain the sustainable flux and recovery of new generation BWRO membranes. Two commercial membranes, selected from bench-scale coupon testing, were evaluated at pilot scale. The results suggested that due to the presence of high silica, operational flux and recovery were limited to 16.9 Lm^− 2 h^− 1 and 75%, respectively, for both RO membranes. The silica fouling predominantly occurred in the second stage of the RO systems, and membrane autopsy revealed that the foulant layer was primarily comprised of inorganic elements (10% organic, 90% inorganic) including silicon and a small amount of aluminium. The antiscalant and chemical cleaning were essential to minimize silica scaling and to restore flux; however, the flux decline rates were faster in the subsequent cycles indicating the detrimental effects of silica deposition. The annual chemical cost could be significantly high due to the increased frequency of chemical cleaning.

Introduction

Brackish groundwater constitutes more than half of the total available groundwater comprising the global water balance [1]. This water is expected to be considered more frequently as an alternative source of drinking water in the near future [2], [3], [4], [6]. Based on a recent survey, about 19% of the desalination plants in the world are installed for brackish water treatment [5]. Brackish groundwater typically contains a high level of total dissolved solids (TDS) ranging from 1000–10,000 milligrams per liter (mg/L). As such, the treatment of this water requires a desalination process such as reverse osmosis, electrodialysis reversal, multiple effect distillation, or multistage flash evaporation. Reverse osmosis (RO) is one of the most widely considered treatment technologies for brackish groundwater desalination in the US and other parts of the world [7], [8], [9], [10]. According to a recent survey, about 59% of the total desalination systems installed in the world (i.e. 13,869 desalination plants) are reverse osmosis systems [5]. Moreover, with the continual emergence of micro-contaminants, water treatment trains are increasingly being designed to achieve multiple treatment objectives that can be effectively met by this technology [10].

High concentration of ions in the brackish groundwater source often increase membrane scaling through precipitation of sparingly soluble salts [11], [12]. In brackish RO systems, the dissolved salts are normally concentrated 4 to10 times, often causing high concentrations of ions exceeding the solubility at the membrane surfaces [13], [14]. Therefore, the inorganic composition and scaling potential of the brackish groundwater need to be carefully examined prior to designing RO systems for desalination. While seawater chemistry is dominated by sodium and chloride ions; divalent calcium and sulfate ions constitute a major fraction of the TDS content of the brackish groundwater [15], [16]. Thus, mineral scaling primarily through crystallization and particle formation at the membrane surface by these ions are important factors determining the operational conditions of the RO process [17], [18], [19], [20]. Calcium sulfate, calcium carbonate, and barium sulfate are some of the most common inorganic salts responsible for surface (i.e., heterogeneous) crystallization and scaling on the membrane surface [21], [22]. In order to alleviate mineral precipitation and improve membrane productivity, strategies like pH adjustment, antiscalant addition, and chemical cleaning have been successfully used [23], [24], [25], [26].

The presence of high levels of silica along with sparingly soluble ions may pose substantial challenge to the implementation of brackish groundwater desalination using RO membranes [27]. The complex chemistry of silica could cause scaling effects in various ways [28]. A number of mechanisms of silica scaling have been hypothesized including (i) colloidal submicron silica deposition on the membrane in an agglomerated form [29], (ii) polymerization of silicates in the bulk solution and on the membrane surface [30], [31], and (iii) co-precipitation of silicates and deposition on the membrane surface [32]. Koo et al. (2001) showed that an increase in the concentration of calcium and magnesium enhanced the polymerization of silica on the membrane surface [33].

Many of these published studies investigated silica scaling on RO membranes through short-term bench-scale tests, some of which demonstrated that the scaling caused by silica could be different compared to mineral salt precipitation. For example, Braun et al. demonstrated that the scaling layer formed by divalent ions (e.g., Ca²⁺) is significantly different from the scaling layer developed by silica [34]. Forensic evaluation of the exhausted membrane surface indicated that silica formed an amorphous solid layer rather than a crystalline solid typically caused by heterogeneous crystallization of divalent ions [29], [33]. Therefore, due to the difference in scaling mechanisms, the effectiveness of mineral scaling inhibition strategies (e.g., antiscalant addition, chemical cleaning) in minimizing the effect of silica scaling on the membrane performance is still requires better understanding. For instance, there is still a need to understand the impact of silica on the RO membranes over an extended period of time in order to determine the impact of silica scaling on the chemical cleaning interval and consequently on annual chemical consumptions.

Furthermore, bench-scale testing and computer modeling approaches for membrane performance prediction have inherent limitations to evaluate the effects of antiscalants and chemical cleaning on the sustainable flux and recovery. Such information could be critically important for widespread application of the RO for the brackish groundwater treatment containing high concentrations of silica. The specific objectives were to: (a) evaluate the long-term effect of high silica on the sustainable flux and recovery of two BWRO membranes, (b) understand the rate of flux decline for a number of subsequent cleaning cycles, (c) determine the effectiveness of antiscalants and chemical cleaning minimizing the impacts of silica scaling, and (d) estimate economic impacts of the silica scaling on the chemical consumption.