A pilot scale pelletiser, after a special start-up with 3,000 mg/l kaolin, was tested continuously for 72-84 h for each run with water from a local river, the Chao Phraya, for its turbidity removal efficiency at different upflow velocities, i.e. (A) 9.6 and 15 m/h for the 'low' turbidity feed water of 30-60 NTU in the summer season, and (B) 6 and 9.6 m/h for the 'high' turbidity influent of 100-200 NTU in the rainy season. Alum, in doses of 3-10 mg/l for Case A and 19-26 mg/l for Case B, and 0.3 mg/l nonionic polymer, were used as the coagulant and coagulant aid for the two cases, respectively. Alternatively, the polymer, at doses of 0.1-0.3 mg/l, was used as the sole coagulant in the process.
The aluminium to turbidity (Al/T) ratios were 0.004-0.015 and 0.014-0.022 g/g SS for Cases A and B, respectively. The polymer to turbidity ratio (P/T) was set at 0.0054-0.0059 and 0.0020-0.0030 g/g SS for the two cases, respectively. The pellet mass loading, in terms of the turbidity removal per pellet mass, was 0.40-0.76 and 0.13-0.21 (g SS/day)/g, or 0.90-1.21 and 0.13-0.20 raw water NTU/g, for the two cases respectively. The pellet retention time (PRT) or pellet age (thetac), required to produce water of better than 5 NTU quality, was 4.4-11.9 and 1.3-3.7 days for the two scenarios. The nonionic polymer was used as the destabilising coagulant with less satisfactory results. The PRT under this condition was, however, 4.8-11.9 and 2-3 days for Cases A and B, respectively. It is noted that the pellet size for both cases was similar, i.e. 0.18-0.22 mm diameter, with a corresponding high density of 1.06-1.12 g/cm3.