| Three coagulants for water treatment: alum, polyaluminum chloride and a low-cost prehydrolyzed coagulant, polyaluminum hydroxysulfate, were evaluated using standard jar tests to determine conditions under which they provided effective coagulation. Various organic species were examined and tannic acid was chosen as a model for natural organic matter. An empirical relationship was developed to describe the sensitivity of each coagulant to organic matter concentration at approximately 5°C and 22°C. Alum was shown to be a poor coagulant in cold water, but to possess a similar sensitivity to organic matter at both temperatures. The prehydrolyzed coagulants were more efficient than alum in cold water, but both displayed an increased sensitivity to organic matter in warm water.; In order to establish some of the reasons behind the differences in coagulant efficiency, the size of the flocs formed was measured using image analysis, and the speciation of aluminum was analyzed using 27Al nuclear magnetic resonance spectroscopy. The prehydrolyzed coagulants were found to produce larger flocs than alum. Smaller flocs were formed by both alum and polyaluminum chloride in cold water, but polyaluminum hydroxysulfate produced particles that were similar in size at both temperatures. The presence of organic matter at low concentrations resulted in the production of larger flocs, but higher concentrations caused a decrease in particle size. From the 27Al nuclear magnetic resonance spectra of coagulant solutions, it was shown that alum contained only mononuclear species and an alumino-sulfate complex; the prehydrolyzed coagulants contained polynuclear species, as well. The relative proportion of the tridecamer species increased on dilution, but decreased on the addition of organic matter. The results of the jar tests and nuclear magnetic resonance analyses indicated a negative correlation between the amount of tridecamer and a coagulant's ability to remove organic matter. |
