Abstract
Laboratory column studies were conducted at the Utah Water Research Laboratory, Logan, Utah, to evaluate reovirus removal from drinking water supplies by slow-rate sand filtration (SSF). Columns, constructed to simulate a full-scale SSF field operation, were inoculated with reovirus at ca. 1,000-times-greater concentrations than those typically found in domestic sewage. Reovirus removal and inactivation were investigated as functions of filter maturity and other filter sand characteristics. Reovirus removal studies demonstrated that the SSF process is capable of reducing reovirus in influent water by a minimum of 4 log concentration units under certain conditions of water quality, flow rate, and sand bed construction. Infectious reovirus was not detected in effluent samples from any of the sand beds studied, after inoculation of the SSF columns; therefore, removal efficiencies were not affected significantly by characteristics, including age, of the two filter sands evaluated. Studies conducted with radioactively labeled reovirus demonstrated that reovirus removed from influent water was distributed throughout the entire length of the filter beds. Concentrations of reovirus in the filter sands decreased with increasing bed depth. The greatest removal occurred in the top few centimeters of all sand beds. No infectious reovirus could be detected in clean or mature sand bed media, indicating that reoviruses were inactivated in the filter.