The evaluation of various schemes to eliminate Combined Sewer Overflows (CSO) from the City of Toronto to the lower Don River and the shores of Lake Ontario resulted in the selection of a storage tunnel system, from which flows are pumped to a treatment facility before discharge to the receiving waters. This paper discusses the procedure used and the results obtained to optimize the combination of the storage volume and treatment rate to meet the required objectives.

To analyze the effectiveness of this scheme in reducing the number of overflow events and the total volume of the CSO, long term simulations were carried out. The City's Quantity-Quality Simulation model (QQS model, developed by Dorsch Consult) was calibrated and run for four years of rainfall record. These years were selected as two average precipitation years, a dry year and a wet year representing the last twenty years of precipitation record.

The model was used to predict the frequency of overflows resulting from different storage volumes (represented by tunnel sizes) and different treatment rates (represented by pumping rates from the tunnel). A family of curves was plotted using the QQS results, to present the relationship between the storage volume and the treatment rate vs. the effectiveness of the scheme presented by the resulting number of CSO events per year, and the percentage reduction in the overall annual CSO volume. To optimize the scheme the total cost of the storage and treatment was estimated, and a family of curves was produced among alternative schemes consisting of combinations of storage sizes and treatment rates, their effectiveness and their overall cost. The curves were plotted to present the effectiveness in the entire waterfront area, as well as the effectiveness in separate areas, since different water quality objectives were set up according to the water use activities in those areas. The procedure then identified the most cost effective combination of tunnel sizes and treatment capacity to meet the objectives in the different areas.