Hydraulic Behavior of Combined Sewer Overflow

Introduction

  • Combined sewer overflow (CSO) is a significant environmental concern due to its role as a major source of water pollution during rainfall events. CSOs release untreated wastewater into natural water bodies, posing risks to aquatic ecosystems and human health.

  • Effective management of combined sewer systems requires accurate prediction of CSO occurrence to mitigate its adverse impacts. Real-time prediction allows for timely interventions such as storage, treatment, or controlled release of excess wastewater.

  • Simulation models, while accurate, often demand extensive and detailed data inputs that may not be readily available or easily obtainable, particularly for large urban areas. These data include sewer network characteristics, land use details, and high-resolution rainfall data.

  • Rainfall data, conversely, is frequently available for extended periods and at high resolution, making it a valuable resource for characterizing and predicting CSO events. Analyzing rainfall patterns can help identify critical thresholds that trigger CSO events.

  • This study aims to develop a simplified yet effective method for predicting CSO hydraulic behavior by focusing on key rainfall parameters that are critical for CSO occurrence. The goal is to provide a practical tool that can be implemented with readily available data to support CSO management decisions.

Methods

  • The study area is the Shingashi urban drainage area located in Tokyo, Japan, characterized by a combined sewer system with 67 CSO outfalls. This area faces challenges related to urban runoff and wastewater management due to its dense population and extensive impervious surfaces.

  • Rainfall data from the year 2007, recorded at a high temporal resolution of 1-minute intervals and a spatial resolution of 1-mm, were utilized. This high-resolution data allows for precise analysis of rainfall intensity and duration, crucial for CSO prediction.

  • Discrete rainfall events were defined based on a minimum inter-event dry period of 4 hours without rainfall. This criterion resulted in the identification of 117 distinct rainfall events, ensuring that each event was independent for analysis.

  • InfoWorks CS software, a widely used hydraulic modeling tool, was employed to simulate rainfall runoff and water flow within the sewer network. The software accounts for various factors such as pipe geometry, surface runoff coefficients, and flow routing algorithms.

  • The study incorporated real, simulated, and calculated data to validate the proposed method. Simulation outputs from InfoWorks CS were considered representative of real-world data, providing a basis for comparison with calculated CSO parameters.

Identification of CSO Occurrence Threshold

  • A series of simulations were conducted for each of the 117 rainfall events, recording CSO occurrence at each of the 67 outfalls. This comprehensive dataset formed the basis for identifying rainfall intensity thresholds for CSO events.

  • A threshold value of rainfall intensity that triggers CSO occurrence was identified for each outfall. This threshold represents the minimum rainfall intensity required to exceed the sewer system's capacity, leading to CSO events.

  • The CSO occurrence threshold was determined by maximizing the coincidence rate, which measures the accuracy of predicting CSO occurrence based on the identified threshold:

Coincidence\ rate\ (%) = \frac{Number\ of\ events\ correctly\ judged}{Total\ events\ (n=117)} \times 100

  • Coincidence rates ranged from 85 to 100% across all outfalls, indicating a high level of accuracy in predicting CSO occurrence using the identified rainfall intensity thresholds. These high coincidence rates validate the reliability of the threshold-based approach.

  • CSO occurred at least once in 56 out of the 67 outfalls, indicating the widespread nature of CSO events in the study area. The identified threshold values ranged from 1 to 34 mm/h, reflecting the variability in sewer system capacity and drainage characteristics across different outfalls.

Calculation of CSO Hydraulic Parameters

  • CSO parameters, including response time, duration, and depth, were calculated based on rainfall parameters. These calculations provide insights into the dynamic behavior of CSO events.

  • The equations used for calculating CSO parameters are as follows:

    • CSO response time, Tr = Tb - T_a

    • CSO duration, T_d = \sum t(i) \quad (i = 1, 2, 3, …, n)

    • CSO runoff depth, V = \varphi \sum_{i=1}^{n} \int (Ii - Ic) dt*i

  • The CSO response time indicates the rapidity with which the sewer network responds to rainfall events, reflecting the time lag between the start of rainfall and the onset of CSO.

  • The CSO duration is calculated as the cumulative time during which rainfall intensity exceeds the established threshold, providing a measure of the length of CSO events.

  • CSO depth is calculated by integrating the excess runoff volume over the duration that rainfall intensity surpasses the threshold, offering an estimate of the total volume of discharged wastewater during CSO events.

Comparison of Calculated and Simulated CSO Hydraulic Parameters

  • A subset of 49 outfalls with more than five CSO events were selected for detailed investigation. This selection criterion ensured sufficient data for robust comparison between calculated and simulated CSO parameters.

  • Calculated and simulated response times exhibited strong correlation, with R2 values of 0.90 and 0.70 for two example outfalls. These high R2 values indicate a close agreement between the calculated and simulated response times, validating the accuracy of the proposed method.

  • CSO duration demonstrated even higher consistency between calculated and simulated results, with R2 values approaching 1.0. This near-perfect correlation suggests that the calculated CSO duration accurately reflects the simulated duration.

  • The distributions of R2 values for response time and duration across all outfalls further illustrate the overall performance of the method.

  • The calculated CSO depth closely matched the simulated depth for all investigated events, with a maximum error of 18%. This small error margin underscores the reliability of the proposed method in estimating CSO volume.

Conclusions

  • A clear threshold for CSO occurrence and hydraulic behavior was successfully identified, offering a straightforward criterion for predicting CSO events based on rainfall intensity.

  • CSO events are triggered only when rainfall intensity exceeds a specific threshold value, providing a practical guideline for CSO prediction and management.

  • CSO response time, duration, and depth were accurately predicted using the proposed method, demonstrating its effectiveness in characterizing CSO hydraulic behavior.

  • The method offers an efficient means of predicting CSO behavior across extensive urban drainage networks, facilitating timely interventions and informed decision-making.

  • The identification of the threshold intensity enhances understanding of the intricate relationship between rainfall patterns and CSO events, contributing to improved CSO management strategies.