Abstract
This study investigated the sustainable reuse of wastewater using fertilizer drawn forward osmosis (FDFO) process through osmotic dilution of commercial nutrient solution for hydroponics, a widely used technique for growing plants without soil. Results from the bench-scale experiments showed that the commercial hydroponic nutrient solution (i.e. solution containing water and essential nutrients) exhibited similar performance (i.e., water flux and reverse salt flux) to other inorganic draw solutions when treating synthetic wastewater. The use of hydroponic solution is highly advantageous since it provides all the required macro- (i.e., N, P and K) and micronutrients (i.e., Ca, Mg, S, Mn, B, Zn and Mo) in a single balanced solution and can therefore be used directly after dilution without the need to add any elements. After long-term operation (i.e. up to 75% water recovery), different physical cleaning methods were tested and results showed that hydraulic flushing can effectively restore up to 75% of the initial water flux while osmotic backwashing was able to restore the initial water flux by more than 95%; illustrating the low-fouling potential of the FDFO process. Pilot-scale studies demonstrated that the FDFO process is able to produce the required nutrient concentration and final water quality (i.e., pH and conductivity) suitable for hydroponic applications. Coupling FDFO with pressure assisted osmosis (PAO) in the later stages could help in saving operational costs (i.e., energy and membrane replacement costs). Finally, the test application of nutrient solution produced by the pilot FDFO process to hydroponic lettuce showed similar growth pattern as the control without any signs of nutrient deficiency.
Original language | English (US) |
---|---|
Pages (from-to) | 18-28 |
Number of pages | 11 |
Journal | Separation and Purification Technology |
Volume | 181 |
DOIs | |
State | Published - Mar 10 2017 |
Bibliographical note
KAUST Repository Item: Exported on 2020-10-01Acknowledgements: The research reported in this paper is part of a collaborative project between the King Abdullah University of Science and Technology (KAUST) and the University of Technology Sydney (UTS) and funded through a SEED Fund provided by KAUST. Support was also provided to HKS by the Australian Research Council (ARC) through Future Fellowship (FT140101208).