Methods

The raw water source was the Moorabool River near Anakie, Australia. This water has a relatively high total organic carbon content (8-10mg/L) and relatively low turbidity (3.9 NTU). The water was treated in standard jar tests with various combinations of magnetite (Fe304), alum (A12(S04)3.18H20), and LT-20 polyelectrolyte (a high molecular weight, non-ionic polyacrylamide). The jar test procedure was to adjust the pH of 1.00kg of raw water from 7.5 to 6.0 with sulphuric acid. The water was then dosed with magnetite, alum, and polyelectrolyte as required. After each chemical addition the water was stirred for 60s at 250RPM. The pH was maintained at 6.0 throughout the procedure with sulphuric acid or sodium hydroxide. Once dosing was complete the mixture was stirred for 90s at 100RPM. The treated water was analysed for dissolved organic carbon (O/I Analytical 1010 Wet Oxidation TOC Analyser). All samples for water quality were filtered through a 0.45pm filter (Selby-Biolab HPLC-certified) prior to analysis. In hybrid magnetite-alum treatment cases, the extent of magnetite-alum attachment was measured from the residual aluminium content in the treated water by inductively-coupled plasma (Jobin-Yvon JY24 ICP spectrometer) after the jar contents were allowed to settle for 150s.

Microfiltration experiments were performed on the single-fibre filtration rig shown in Figure 1. The membranes were polypropylene hollow fibres with a nominal pore size of 0.2pm, an internal diameter of 250 m, an external diameter of 550pm, and a length of 1.2m. Fibres were sealed at the open ends with a silicone septum, wet with ethanol and flushed with pure water. The feed water was pressurised with a peristaltic pump, and forced through the hollow-fibre membrane to emerge as permeate from the open ends of the fibre. The permeate was collected in a vessel mounted on an analytical balance. A pressure transducer measured feed pressure. The signals from the analytical balance (WT) and the pressure transducer (PT) were processed to calculate the permeate flowrate as a function of permeate throughput. The feed water was treated using the standard jar test procedure described above, although the 100RPM stirring regime was omitted. The feed suspension was stirred at 250RPM and the pH was controlled at 6.0 throughout the membrane fouling experiments, which took 2-3 hours. In some cases, the feed suspension was allowed to settle, and only the supernatant was pumped onto the membrane.

0 0

Post a comment