This paper presents the results of pilot-plant scale investigation of the feasibility of removing calcium scale precursor ions from saline municipal wastewater via precipitation, thereby increasing the achievable water recovery. The adopted precipitation process involved the addition of phosphate to generate a seed material which served as a seed for further precipitation. An increase of pH was adopted to decrease the solubility of chemical species to be precipitated, and the precipitated solids were removed by filtration. Laboratory trials in preparation for this pilot plant trial showed that it was possible to reduce the calcium concentration of brine from 250 mg/L to approximately 10 mg/L using approximately 200 mg/L phosphate ion addition at pH 10. The process was tested at the pilot plant scale on municipal wastewater, in Donald, a small Victorian town in the Wimmera Mallee region. Addition of phosphate was required for the removal of calcium scale precursors from the saline wastewater. A high concentration of magnesium prevented precipitation of calcium carbonate and calcium sulphate even if the phosphate ions, a known inhibitor of calcium carbonate and calcium sulphate precipitation, were removed from solution via ferric chloride coagulation. Precipitation of calcium phosphate was able to achieved low calcium concentrations from RO brine (<10 mg/L) for prolonged periods of time.
However, excess phosphate doses were used because short run times prevented optimization of the phosphate dose. The short run times resulted from high turbidity episodes in the filtrate of the ceramic ultrafiltration (CUF) used to remove the precipitated calcium phosphate. These spikes in turbidity were associated with increased temperatures arising from the high recycling rates and cross flow velocities across the ceramic membrane. It is proposed that calcium phosphate precipitated from solution once it had passed through the ceramic membrane, and the lower solubility of calcium phosphate at higher temperatures supports such a mechanism.
During this trial, the RO membranes were prone to fouling from the excess phosphate added. This suggests that optimal control of the phosphate dose would be required to prevent significant RO fouling.