Elsevier

Journal of Membrane Science

Volume 428, 1 February 2013, Pages 181-189
Journal of Membrane Science

Organic fouling in pressure retarded osmosis: Experiments, mechanisms and implications

https://doi.org/10.1016/j.memsci.2012.10.045Get rights and content

Abstract

Pressure retarded osmosis (PRO) is an osmotically-driven membrane process and it has attracted increasing interest for salinity-gradient power harvesting. However, the PRO performance (both water flux and power density) can be significantly limited by membrane fouling. This study, for the first time, systematically investigated membrane fouling by organic foulants in PRO process and its effect on PRO power density. It was found that significant alginate fouling occurred when the draw solution (DS) contained large quantities of divalent cations (i.e., Ca2+ and/or Mg2+). This is attributed to the reverse solute diffusion enhanced organic fouling, the effect of which is related to (1) the type of draw solute and the rate of its diffusion into the feed solution (FS), and (2) its ability to interact with feed foulant. It was also found that the increase of DS concentration exacerbated the PRO fouling due to the synergistic effects of increased reverse solute diffusion and increased initial water flux level. However, the increase of applied hydraulic pressure mitigated the alginate fouling for NaCl draw solution but exacerbated the alginate fouling for CaCl2 draw solution due to the competing effects between the increased reverse solute diffusion and the reduced initial water flux. It was further found that the alginate fouling was more severe under PRO operation than that under forward osmosis (FO) operation at an identical initial water flux level using seawater-based DSs due to the faster reverse solute diffusion under PRO operation. Our results provide significant implications for PRO fouling control.

Highlights

► Organic fouling in pressure retarded osmosis (PRO) was systematically investigated. ► PRO fouling was more severe at higher water flux (Jv). ► Reverse solute diffusion (Js) played an important role in enhancing PRO fouling. ► PRO fouling was more severe at increased draw solution concentration due to higher Jv and Js. ► Increased pressure affected PRO fouling by the competing effects of decreased Jv and increased Js.

Introduction

Pressure retarded osmosis (PRO) has gained growing interest for its potential of harvesting osmotic power from the mixing of two liquid streams with different salinities [1], [2], [3]. It is an osmotically-driven membrane process, where water flows from a low-osmotic-pressure feed solution (FS) to a high-osmotic-pressure draw solution (DS) against a hydraulic pressure [4]. This process converts the osmotic power into a mechanical energy, whose power is equal to the product of water permeation rate and applied hydraulic pressure [4]. The mechanical energy can be subsequently converted to other forms of useful energy (e.g., electricity by running the pressurized DS through a hydroturbine) [4]. Recent research has significantly advanced this technology from process evaluation [5], [6], [7], [8], membrane fabrication [9], [10], and module development [11], [12], [13].

Despite numerous research efforts on PRO, membrane fouling in PRO processes has not been investigated to date. In practice, a variety of low-salinity feed solutions can be used in PRO processes, such as river water, impaired water or pretreated wastewater effluent, and brackish water. In these feed solutions, organic foulants are ubiquitous [14], [15], [16] and can potentially cause severe fouling problems [16], [17], [18], [19], [20], leading to undesirable PRO performance decay in terms of both water flux and power density. To understand the intrinsic cause and effect of organic fouling in PRO processes is of paramount importance for the development of effective PRO fouling-control strategies.

Previous studies of membrane fouling in forward osmosis (FO, a similar process to PRO where there is no applied hydraulic pressure) revealed that organic fouling is affected not only by the feed solution composition [19], [21] but also by the draw solution composition [17], [21], which was attributed to the mechanism of reverse solute diffusion enhanced organic fouling [17], [21]. The solutes reversely diffusing from DS into FS can potentially interact with the foulants in the FS (e.g., Ca2+ interacting with alginate) and thereby enhance organic fouling of FO membranes [17], [21]. Since reverse solute diffusion also occurs in PRO processes [7], [22], this mechanism is likely applicable for PRO fouling although additional studies are needed to verify this assumption. In particular, DSs in PRO processes may include seawater (SW) or waste brine from SW desalination plant [5], [6], [16], [23]. These SW-based DSs contain substantial amount of divalent cations (e.g., Ca2+ and Mg2+) that can act as organic fouling initiators when they reversely diffuse into FS [17]. Moreover, the rate of reverse solute diffusion depends on the applied pressure during PRO operation [7], [8]. Faster reverse solute diffusion at higher applied pressures can potentially lead to more severe organic fouling in PRO processes [17]. On the other hand, the increase of applied pressures can reduce the water flux in PRO processes [7], which may mitigate membrane fouling [17], [19], [21], [24], [25], [26], [27]. This suggests that membrane fouling behavior may become much more complicated under PRO operation, which deserves further attention in the PRO communities.

The objectives of this study are (1) to systematically investigate the effects of draw solution and feed solution composition, draw solution concentration, and applied hydraulic pressure on the PRO fouling, and (2) to explore the mechanisms governing the organic fouling in PRO processes. The results from the current study have important implications for PRO fouling control.

Section snippets

Membranes

A commercial membrane provided by Hydration Technology Innovations (HTI, Albany, OR) was used in this study. It has a cellulose triacetate (CTA) rejection layer and a woven fabric support layer (denoted as CTA-P following our previous study [7]). The membrane has a water permeability of 2.08×10−12 m/s Pa, a NaCl permeability of 1.76×10−7 m/s, and a structure parameter of 0.48 mm [7]. Once received, the membrane samples were stored at 4 °C. They were extensively rinsed and soaked in Milli-Q water for

Effect of draw solution type

Fig. 1 and Fig. C1 (Appendix C) show the effect of draw solution composition on PRO fouling in the AL-DS orientation for an FS containing 100 mg/L alginate in 10 mM NaCl at pH 6.2. Draw solutions investigated include SWBr, 1 M NaCl, 0.7 M CaCl2 and 0.7 M MgCl2, whose osmotic pressures at 25 °C are 49.3, 46.8, 51.9 and 57.9 bar, respectively, to achieve similar initial water flux level (∼17–19 L/m2h). As the experiments were conducted at identical effective hydraulic pressure (6.5 bar), both water flux

Conclusions

The current study for the first time systematically studied the PRO fouling and revealed that reverse solute diffusion can strongly affect PRO fouling, the effect of which is related to (1) the type of draw solute and rate of its reverse diffusion into feed solution, (2) the type of foulant in the FS, and (3) the extent of intermolecular interaction between the foulant and reversely diffused draw solute in the FS.

In addition, water flux level is another important factor affecting PRO fouling,

Acknowledgments

This research grant is supported by the Singapore National Research Foundation under its Environmental & Water Technologies Strategic Research Programme (grant # MEWR C651/06/173) and administered by the Environment & Water Industry Programme Office (EWI) of the PUB. The authors also thank HTI for supplying membrane samples. Mr. Pan Fanfan is thanked for his assistance on PRO experiments.

References (42)

Cited by (158)

View all citing articles on Scopus
View full text