Long-Term CSO Control Plan for the Pittsburgh Water & Sewer Authority (PWSA)

PWSA is currently developing a CSO Long-Term Control Plan (LTCP). The purpose of this paper is to present a summary of the technical process used to develop the LTCP and an overview of its status. This effort included data collection and analysis in the collection system and receiving waters, and development and calibration of a dynamic hydraulic collection system model.An extensive collection system flow monitoring program was implemented. It included 418 temporary flow meters installed for a 4-month period. A sub-set of 19 of these meters was left in place for four additional months. This extensive data set gave a detailed picture of flows throughout PWSA's collection system.Sampling was also conducted on a representative set of six CSO outfall discharges. Discrete samples were taken at each outfall for two to three storm events and analyzed for bacteria and solids. Six major tributaries were also sampled using a 4-parameter (DO, pH, temperature and conductivity) data sonde. These in-stream meters were in place for about three months during the summer. Water quality data for the major rivers (Allegheny, Monongahela and Ohio) were reviewed from ongoing water quality data collection programs by others including USGS.A highly detailed model of PWSA's collection system was developed using the InfoWorks™ model.After validation, the model was used to assess CSO activation volumes and frequencies by simulating a typical year of rainfall. It was also used to assess system capacity and hydraulic limitations by using a series of design rainfall events. A baseline condition was then established by adding projects to the model that were to be completed within the next two to three years.Alternatives were developed for a range of CSO control levels, including 0, 1, 2, 4 and 6 overflows per year. A systematic process was used where a range of alternatives were developed and evaluated, first for each individual outfall, then for geographic groupings of outfalls, then for major systems and system-wide. This allowed for the best solutions to be carried forward, whether they applied to individual outfalls or groupings. Cost performance curves were then constructed using the best combination of solutions for each control level.