Calcium phosphate scaling in osmotically driven membrane processes: Limiting flux behavior and its implications for scaling mitigation

Abstract

Abstract Calcium phosphate scaling, a typical inorganic scaling, has been identified as a challenging issue in the operation of osmotically driven membrane processes (ODMPs) especially in pressure-retarded osmosis (PRO) mode where membrane porous substrate faces against the impaired feed solution (FS). During scaling in PRO mode, it has been observed that flux declines to a pseudo-stable level - limiting flux, after which further flux decline is negligible. However, the underlying mechanisms of limiting flux and its potential impact on practical PRO processes is poorly understood and necessitate systematic exploration. Herein, in this study, the behavior of limiting flux was examined during calcium phosphate scaling under various solution conditions and the associated mechanisms were elaborated. Experimental results show that for a given membrane and a given water source (a fixed FS composition), an increase in the concentration of the draw solution (DS) has an insignificant impact on the final limiting flux. In contrast, an increase in the concentration of dominant scaling precursor ions (i.e., Ca2+ and PO43− in this study) in the FS significantly decreases the stabilized limiting flux. Therefore, scaling mitigation via adjusting FS pH and adding chelating agent EDTA was investigated. This was demonstrated to be effective in retarding flux decline by lowering the concentration of dominant scaling precursors, leading to an increase in the stabilized limiting flux. Physical cleaning including osmotic backwash (OBW) and surface flushing after scaling was also conducted and compared. While OBW is generally more effective than surface flushing to restore the water flux, it is less effective for scaling dominated by surface crystallization compared to that by bulk crystallization. Finally, we emphasize the significance of limiting flux for optimizing practical PRO performance and we identify other knowledge gaps that should be investigated to further improve PRO performance.