Application of ultrafiltration membranes for removal of humic acid from drinking water

Abstract

Humic acids are primarily a result of the microbiological degradation of surrounding vegetation and animal decay and enter surface waters through rain water run-off from the surrounding land. This often gives rise to large seasonal variations, high concentrations in the wet season and lower concentrations in the dry season. Alone humic acid is just a colour problem but when present in conventional treatment processes like chlorination, carcinogenic by-products like trihalomethane and haloacetic acid are formed. This, in addition to the demand for clean potable drinking water, has sparked extensive research into alternative processes for the production of drinking water from various natural/industrial sources. One of the major areas of focus in these studies is the use of membranes in microfiltration, ultrafiltration and nanofiltration. In this report the humic acid removal efficiency of ultrafiltration membranes with 3 kDa, 5 kDa and 10 kDa MWCO is examined. The membranes were made of regenerated cellulose and were in the form of cassette providing a 0.1 m 2 surface area. At first distilled and deionised water, known as milliQ water, was used as the background feed solution to which humic acid powder was added. It was found that all three membranes removed humic acid with an efficiency of approx. 90% and were capable of reducing initial concentrations of 15mg/L to below the New Zealand regulatory limit of 1.17 mg/L. The permeate flux at a transmembrane pressure of 2.1 bar was approx. 20 l/m 2/h (LMH) and 40 LMH, respectively through the membranes with MWCO 3 kDa and 5 kDa. These membranes experienced significant surface fouling resulting in retentate flow rates as low as 11 litres per hour after just four runs compared to the recommended 60–90 l/h. Cleaning with 0.1 M NaOH slightly improved the retentate flow rate, but well below those obtained with fresh membranes. The 10 kDa membrane provided high retentate flow rates which evidently minimised fouling by providing a good sweeping action across the membrane surface while maintaining humic acid removal below the regulatory 1.17 mg/L level. The permeate flux through this membrane was initially high (140–180 LMH) and reduced to approx. 100 LMH after 10–12 min of operation. Increasing the initial humic acid feed concentration from 10 mg/L to 50 mg/L did not significantly decrease humic acid removal efficiency although the retentate flow rate was lower at higher concentrations. Finally the tap water was tested as the background solution and treated for the removal of humic acid. The presence of ions and other impurities in the tap water had little effect on humic acid removal. However, the permeate flux through 10 kDa membrane decreased from 100 LMH for milliQ water to 60 LMH for tap water after 20 min of operation.