The objective of this pilot study was to evaluate the performance of Powdered Activated Carbon (PAC) for the removal of individual trihalomethanes (THMs) from chlorinated drinking water, Athens, Greece. The pilot treatment facility was located at the Water Treatment Plant of EYDAP (Athens Water Supply and Sewerage Company) in Galatsi, Athens, and consisted of two mixing tanks operating as open batch reactors, which were fed with chlorinated water from the overflow of the sedimentation tanks. Experimental cycles were performed, with different PAC doses ranging from 5 to 50 mg L-1 and with contact times 30 and 60 min for each cycle. Water samples taken at the start of each cycle and after contact with PAC, were analysed for individual THMs and Dissolved Organic Carbon (DOC). Bromide, free residual chlorine, pH and temperature were also measured. From the experimental results, it was shown that all THMs concentrations were decreased after PAC contact, the decrease being higher for higher PAC dose and higher contact time, with few exceptions. This fact probably indicates that adsorption rate plus volatilisation rate of THMs overcame formation rate, in most cases. It was also concluded that volatilisation and possible formation reactions of THMs had probably a more apparent effect to the total removal of most THMs at low PAC doses and low contact time. At higher PAC doses and higher contact time, PAC adsorption probably had an increased contribution to THMs removal, because of the larger carbon surface area and the longer contact time, resulting in higher adsorbed mass according to adsorption theory. By using isotherm results of individual THMs on activated carbon by other researchers, the theoretically adsorbed mass of THMs per g of PAC was roughly estimated and compared with the measured removed mass of THMs per g of PAC at PAC dose 50 mg L-1 and contact time 60 min for most THMs, on the assumption of equilibrium achievement. The removed mass of THMs by PAC appeared much higher than the theoretically adsorbed, especially for the most volatile members. The difference was attributed to volatilisation taking place in the open mixing tanks. In conclusion, the practical results from the application of PAC for the removal of THMs are much better than expected from single adsorption by PAC. Volatility of THMs, especially of chloroform (TCM), substitutes for weak adsorbability and gives increased removal. The influent DOC was almost constant in all experiments. The removal of DOC was generally higher for increased PAC dose and contact time. Although equilibration cannot have happened for DOC, the mass removal of DOC per g of PAC was higher by more than two orders of size than that of the smaller and in lower concentration THMs, as expected.
