Secondary treated domestic wastewater in reverse electrodialysis: What is the best pre-treatment?

Highlights

• RED-RO is a promising technology for energy-efficient seawater desalination.

• Secondary-treated wastewater is a potential low salinity source for RED.

• An efficient pre-treatment of wastewater is needed before use in RED.

• Rapid sand filtration and 100 µm provide efficient wastewater pre-treatment.

• River bank filtration does not provide adequate feed water quality.

Abstract

Although Reverse Electrodialysis (RED) is most commonly known as a selective separation technology used for the production of sustainable energy, it can also serve as a valuable pre-desalination tool. By coupling RED to Reverse Osmosis (RO) for seawater desalination: (1) sustainable energy is produced in the RED process and (2) seawater is partially desalinated prior to RO thus, decreasing the energy demand. In this study, secondary-treated wastewater is proposed as the low salinity source in RED and suitable pre-treatment techniques for this effluent are investigated. Although it is generally accepted that RED is less prone to fouling than typical pressure driven membrane processes, results showed that pre-treatment is a key to ensure efficient operation of the wastewater-seawater RED. Both 100 µm filtration and rapid sand filtration proved to be suitable, with an increase in pressure drop of only 0.09–0.18 bar and a permselectivity decrease of only approximately 20% during 40 days of continuous operation. Conversely, River bank filtration did not perform better than the non-pretreated sample. As such, 100 µm filtration and rapid sand filtration are considered suitable, robust, and cost efficient pre-treatment options for wastewater fed RED, enabling the improvement of the hybrid process of RED-RO seawater desalination.

Introduction

Reverse electrodialysis (RED) is an electrochemical membrane technology generally used for the sustainable production of energy from salinity gradients between high salinity (e.g. seawater or brines) and low salinity streams (e.g. river water) [1], [2], [3]. This technology consists of alternating anion and cation exchange membranes (AEM and CEM), over which a potential difference is created by the salinity gradient between the feed streams, and has been extensively discussed in literature [4], [5]. Recently, a new application of RED, in an RED-reverse osmosis (RO) hybrid, has been suggested [6]. RO is the most widely used technology for drinking water production from seawater. With a typical energy demand of 2–4 kWh/m³ (at typical seawater salt concentrations of 35 g/L and a recovery of 50%) [7], [8], [9] it is an energy intensive process, limiting its applicability in developing regions facing water shortages. Options for decreasing the energy demand of RO are limited, as the thermodynamic limit is 1.06 kWh/m³ [7], [10]. By coupling RED to RO the energy demand can be decreased by: (a) producing energy out of the salinity gradient between the RO feed or brine and a low salinity stream and, (b) decreasing the seawater salt concentration prior to RO. This RED-RO hybrid system was extensively discussed in Vanoppen et al. (2016) [11].

However, the choice of the low salinity stream for this system is critical, as it does not make sense to use fresh water suitable for drinking water production for this purpose. The use of impaired water (e.g. secondary treated wastewater) is therefore suggested, as this water is generally not used for drinking water production due to legislative, social or cultural barriers. The presence of microcontaminants in impaired also requires additional treatment, increasing energy demand. Combining RED and RO would provide a double barrier, preventing micropollutants in the impaired water from ending up in the final product water.

Although it is generally accepted that electrochemical membrane systems are less prone to fouling compared to pressure driven membrane systems [4], [12], [13], an adequate pre-treatment would be required when using impaired water. The most significant fouling in the RED systems used in this study would be of organic and biological nature, because of the use of secondary treated wastewater [13], [14], [15], [16], [17], [18]. The risk for scaling is low, since none of the streams are concentrated beyond the initial concentration of the high salinity stream. However, Vermaas et al. (2013) did observe some scaling on the CEM on the seawater side, resulting from the exchange of multivalent ions from river water across that membrane [13]. In a more recent study, Kingsbury et al. (2017) compared the fouling behavior of different natural water pairs in RED. They concluded that natural organic matter (NOM) plays a major role in RED fouling effects, with NOM decreasing the power density up to 43% [15]. Di Salvo et al. (2018) investigated the use of high salinity fishery wastewater and low salinity wastewater after a membrane bioreactor in RED. They observed clear obstruction of the feed channels and adsorption of foulants onto the AEM [19]. Different anti-fouling strategies are reported. Air sparging is an effective way to combat fouling when using real seawater and river water (after pre-treatment by 10 µm filtration) [20] and Moreno et al. (2017) reported the use of CO₂-saturated water as a two-phase flow for successful cleaning and fouling control [21]. Comparable to electrodialysis (ED) reversal, short ED pulses can be used in RED to detach foulants from the membranes [19]. Nevertheless, the literature on fouling in RED remains limited and has mainly focused on stack design [22], [23], membrane development [24], [25], [26] and cleaning strategies [20], [21] to combat fouling. Fouling and water pre-treatment strategies have not received much attention, especially when considering the use of wastewater as the fresh water source in RED.

The selection of a pre-treatment strategy for the impaired water is critical. To allow application of the technology in rural and developing areas, the pre-treatment technology should be simple, robust and cheap. These were the main selection criteria for the technologies tested in this study; none of them requires additional chemicals or extensive process knowledge to operate. Rapid sand filtration (RSF) is a pre-treatment step commonly used in large scale seawater reverse osmosis applications, due to its simplicity, low energy consumption and low operational costs [27]. It can remove suspended solids as small as 0.35 mm by size exclusion and adsorption and can also serve as a substrate to develop a functional microbial biofilm to remove biodegradable compounds and nutrients from the water [27], [28]. 100 µm filtration is a conventional membrane filtration, capable of retaining bigger particles (e.g. sand, algae, fibers and debris). The buildup of retained particles on the filter can further increase its efficiency because of secondary filtration. In both RSF and 100 µm filtration, salts, humic acids and microorganisms are not efficiently retained. The costs for both is also similar; ±50 000 EUR for an installation that produces 50 m³/day, with operational costs of 0.1–0.15 EUR/m³ for MF and negligible operational costs for RSF [29], [30]. Another option, often used in drinking water production and groundwater recharge and more recently for the removal of micropollutants, is river bank filtration (RBF) [31], [32]. In contrast to RSF, flow rates in RBF are significantly lower, while the size distribution of sand particles is considerably wider. Besides simple filtration, the main removal mechanisms are adsorption and biodegradation of solutes by the biologically active medium, thus effectively removing part of the microorganisms and NOM, contrary to the first two techniques [31]. Costs for RBF depend highly on local conditions.

The goal of this study is to identify potential pre-treatment demands for impaired water (i.e. wastewater treatment plant (WWTP) effluent) to be used in RED and to compare different pre-treatment strategies for impaired water and their effect on RED operations. Three different pre-treatment techniques were selected: RSF, 100 µm filtration and RBF, and their performances in terms of maintaining stable RED operation were compared to RED efficiency without pre-treatment. By combining operational parameters (pressure drop, desalination, …) and microscopic analysis (scanning electron microscopy, atomic force microscopy, …), a comprehensive correlation between fouling and pre-treatment type is obtained. This study provides insights into simple, robust and inexpensive treatment methods and their ability to ensure stable long-term RED operation, which is valuable information considering any type of application using ion-exchange membranes with natural water.