ABSTRACT
Experiments to determine the removal of viruses in different types of water (surface water from two reservoirs for drinking water treatment, treated groundwater and groundwater contaminated with either 5 or 30 % of wastewater) by ultrafiltration were performed with a semi-technical ultrafiltration unit. Concentrations of human adenoviruses (HAdVs), murine norovirus (MNV), and the bacteriophages MS2, ΦX174 and PRD1 were measured in the feed water and the filtrate, and log removal values were calculated. Bacteria added to the feed water were not detected in the filtrates. In contrast, in most cases viruses and bacteriophages were still present in the filtrates: log removal values were in the range of 1.4-6.3 depending on virus sizes and water qualities. Best removals were observed with bacteriophage PRD1 and HAdVs, followed by MNV and phages MS2 and ΦX174. Virus size, however, was not the only criterion for efficient removal. In diluted wastewater as compared to drinking water and uncontaminated environmental waters, virus removal was clearly higher for all viruses, most likely due to higher membrane fouling. For quality assessment purposes of membrane filtration efficiencies with regard to the elimination of human viruses the small bacteriophages MS2 and ΦX174 should be used as conservative viral indicators.
ABSTRACT
Bacteriophages are often used as surrogates for enteric viruses in spiking experiments to determine the efficiencies of virus removal of certain water treatment measures, like e.g. flocculation or filtration steps. Such spiking experiments with bacteriophages are indispensable if the natural virus concentrations in the raw water of water treatment plants are too low to allow the determination of elimination levels over several orders of magnitude. In order to obtain reliable results from such spiking tests, it is essential that bacteriophages behave comparable to viruses and remain stable during the experiments. To test this, the influence of flocculation parameters on the bacteriophages MS2, Qß and phiX174 was examined. Notably, the F-specific phages MS2 and Qß were found to be inactivated in flocculation processes with polyaluminum chloride (PACl). In contrast, other aluminum coagulants like AlCl3 or Al2(SO4)3 did not show a comparable effect on MS2 in this study. In experiments testing the influence of different PACl species on MS2 and Qß inactivation during flocculation, it could be shown that cationic dissolved PACl species (Al13) interacted with the MS2 surface and hereby reduced the surviving phage fraction to c/c0 values below 1*10(-4) even at very low PACl concentrations of 7 µmol Al/L. Other inactivation mechanisms like the irreversible adsorption of phages to the floc structure or the damage of phage surfaces due to entrapment into the floc during coagulation and floc formation do not seem to contribute to the low surviving fraction found for both F-specific bacteriophages. Furthermore, no influence of phage agglomeration or pH drops during the flocculation process on phage inactivation could be observed. The somatic coliphage phiX174 in contrast did not show sensitivity to chemical stress and in accordance only slight interaction between Al13 and the phage surface was observed. Consequently, F-specific phages like MS2 should not be used as surrogate for viruses in flocculation experiments with PACl to determine the removal rates of viruses, as the results are influenced by a strong inactivation of the bacteriophages due to the experimental conditions.