Abstract
In this study the design of a draw solution with optimum osmotic pressure and water flux performance based on a Forward Osmosis process (FO) is investigated. A protocol was used to determine the most appropriate draw solutions to be tested on an FO pilot plant using a hollow fibre membrane. The criteria used for the selection besides solubility are, operational osmotic range, toxicity, flammability and cost. The possible draw solutions (DS) were tested by examining the osmotic behaviour of binary and ternary systems made from electrolyte and non-electrolyte solutes (NaCl, CaCl2, MgCl2, MgSO4, Acetamide, Fructose, Maltose, Sucrose). Their osmotic properties were determined from a series of measurements of water activities using the hygrometric method for solutions of molality ranging from 0. 5 to 6. 0 mol kg-1 at 298. 15 K. The osmotic pressures thus determined were compared with values obtained from van’t Hoff limiting law and OLI software package. The results were in good agreement generally with OLI but not, as expected with van’t Hoff law, which is only applicable at very low concentrations. In general electrolytes performed better than non-electrolytes, where MgCl2 produced the highest osmotic pressure. Positive and negative synergistic effects on osmotic pressure in the ternary system were observed. It was found that all ternary systems made from electrolytes showed positive synergy. But those containing acetamide and electrolytes (except NaCl) showed negative synergy. While the systems containing sugars showed three types of behaviours that are, positive, negative and additive. Besides, ion-pair formation, solute-solute and solute solvent effects as well as possible complexations are all expected in the ternary systems. By following the protocol two ternary systems, MgCl2 + NaCl + H2O and sucrose + NaCl + H2O and their binary counterparts, NaCl, MgCl2 and sucrose were selected to be best suited for the FO application. These solutions were then tested on an in-house FO pilot plant to evaluate the water flux, water recovery and reverse salt diffusion. Their performance was limited by internal concentration polarisation (ICP) significantly reducing the permeate flux. The highest water flux and recovery values were achieved with NaCl binary solution. However, the reverse salt diffusion of NaCl in both ternary DS’s was found to be lower than in its binary form. The study showed that changing the mole ratios of the solutes (MgCl2 +NaCl and sucrose + NaCl) in the ternary DS affected all three performance parameters. The study found that although a DS’s ability to achieve a high osmotic driving force is important in the selection of an optimum DS, high flux performance is vital in designing an optimum draw agent. Key factors found to contribute to FO performance include solute type, ratio of solutes present in a ternary DS, solute-membrane interaction and viscosity of a DS. The different behaviours of these draw solutions highlight the potential of multi-component draw solutions have in designing an optimum osmotic agent for specific FO applications and membrane type.