Development of Boundary Condition Timeseries for Combined Sewer Overflow Collection System Models for a ”Typical Year” to Enhance Model Capabilities

The Philadelphia Water Department (PWD) maintains hydrologic and hydraulic models of the City’s combined sewer collection system for planning, management and compliance purposes. PWD relies on these models to evaluate the effectiveness of existing and proposed combined sewer overflow (CSO) control measures. These models are the foundation of Green City, Clean Waters, Philadelphia’s plan to reduce stormwater entering the combined sewer system through the implementation of green infrastructure, and thus reduce wet weather overflows.The receiving waters in the Philadelphia area include the tidally influenced Delaware and Schuylkill rivers and both non-tidal and tidal extents of the main tributaries Tacony/Frankford Creek, Cobbs Creek, and Pennypack Creek.The Green City, Clean Waters program aims to reduce wet-weather overflows in the Philadelphia combined sewer collection system using the collection system model evaluated on a typical year basis as an analysis tool. The typical year modeling approach was developed to evaluate the effectiveness of current and proposed CSO control measures under average annual conditions. The typical year represents average annual hydrologic conditions through the development of rainfall input timeseries to represent average, or typical, annual input to drive the models. Typical year rainfall was developed based on calendar year 2005 rainfall with modifications to selected events. In similar fashion, it was recently proposed that incorporating tidal boundary conditions that represent average annual dynamic tides would improve the model, and a typical year tide timeseries was developed. An analysis of the water surface levels at the model boundaries was used to develop these boundary conditions for both tidal and non-tidal reaches.This paper will present how the typical year boundary conditions were developed for the collection system models, including rainfall and tidal and non-tidal outfall boundaries of the receiving waters. Methods to develop a representative 12-month rainfall record, or typical year, will be explored, as well as the development of a corresponding typical year tidal timeseries to improve the boundary conditions of the models. Frequency analysis of observed water levels and analysis of empirical cumulative distribution function of long-term data were used to develop the tidal timeseries. Transforming the tidal observations over a region of 5 classified tidal zones was also explored. Future enhancements to the typical year non-tidal boundary conditions are also discussed.