An Incremental, Balanced Approach to Enhanced Wet-Weather Management and Operational Optimization

The Metropolitan St. Louis Sewer District (MSD) is 1 year into their multi-year program to develop a real-time decision support system (RT-DSS) that will manage and operate their wet-weather infrastructure to improve water quality, more effectively use existing infrastructure and control options, and reduce the need for expensive infrastructure to meet MSD's Long-Term Control requirements. MSD envisions an enhanced way of managing its complex wet weather infrastructure that leverages data and technology to augment and elevate staff capabilities and achieve an integrated, coordinated control of wet weather events. This program signals a large transformation for MSD both organizationally and technically. A defined plan, as shown in Figure 1, emphasizing incremental progress has been a touchstone throughout the program in order to demonstrate the concepts, value and challenges of real-time control (RTC) facilities and build buy-in from all stakeholders. The following abstract will discuss the progress to date, including the completion of the existing system's modeling and optimization to identify effective operational strategies to reduce combined sewer overflow (CSO) and the ongoing pilot study to test the potential outcomes of real-time operational control changes. These foundational activities are essential to the future successful deployment of the RT-DSS.
Background
MSD serves 1.3M customers over 520 square miles and has the 4th largest sewer system in the United States. In 2011, MSD entered into a consent order with the US EPA requiring MSD to spend a minimum of $4.7 billion over the next 23 years to address the issue of overflows and other sewer system improvements. Lemay, one of the 5 MSD service area s shown in Figure 2 covers 123 square miles of combined and separate sewered area and was selected as the initial proof of concept for the program. Lemay has 118 CSOs with a total 6,635 million gallons (MG) during the typical year. Most existing CSOs include an Interceptor/Outfall (I/O) gate that controls flow to the interceptor and treatment plant. Under baseline conditions the I/O gates close to 25% during wet-weather, forcing any additional flow to CSO.
Existing System Optimization
The first step of the program was to identify the locations and type of enhanced control strategies in the existing system that provide the highest CSO capture while meeting MSD's hydraulic constraints to avoid interceptor surcharge. Due to the high number of options to operate Lemay's system, an optimization approach was undertaken to intelligently search through the broad range of improvement options. Optimization is a structured process for evaluating the costs and benefits of many alternatives, comparing tradeoffs, and identifying high-performing solutions that function well in coordination. Figure 3 provides an overview of the optimization workflow performed.
Input Development Three major inputs are required for an optimization – the hydraulic model, the project alternatives to consider and criteria to judge alternative performance on. The Lemay combined system hydraulic model was converted to PC-SWMM and updated to include the contributing sanitary system.
The project alternatives were generated during a day-long workshop with MSD staff and included alternative operational strategies at I/O gates, weir height increases and controlling gates at major sanitary infrastructure during wet weather. Performance was measured by the combined cost of these project alternatives, as well as their corresponding CSO reduction and any additional surcharge risk to the interceptor.
Optimization Execution Optimatics' Optimizer software was used to evaluate over 200,000 distinct alternative combinations. Optimizer links to the hydraulic model and leverages cloud computation and an evolutionary algorithm to systematically search for better and better solutions to meet the defined performance objectives.
Results Review The optimization output was reviewed to understand how the system responds to different alternative combinations and which alternative strategies were essential to high performance. Figure 4 is an example of the optimization output illustrating CSO Volume versus Interceptor Surcharge Risk, color coded based on the number of alternatives included in each scenario. The optimization analysis demonstrates there are many ways to operate MSD's existing system to reduce CSO volume by using in-system storage more effectively than the baseline condition. As called out in Figure 4, an over 3% reduction in CSO volume is possible by maintaining the same interceptor surcharge risk as the baseline, which translates into over 200 MG of typical year CSO. Depending on the level of interceptor surcharge acceptable to MSD, the CSO volume reduction could be up to 9% using the existing system, or 470 MG typical year CSO volume. RTC Pilot Study The pilot study serves as an opportunity to test out different operational strategies identified as high-performing from the optimization analysis and provide confidence the CSO benefits translate from the model to the real world. In addition, the pilot study enables MSD to identify and overcome any technical and organizational issues with implementing RTC. Lessons learned from the Pilot Study will inform the future system-wide implementation of RTC. The Rock Creek CSO, a relatively large CSO in Lemay's system, was selected as the Pilot Study due to potential CSO gains by modifying the I/O gate logic to remain open longer during wet-weather events and maximize use of the existing system storage. A representation of the Rock Creek CSO diversion structure is shown in Figure 5. In order to track the Pilot's performance, an online dashboard was built that integrates a variety of data streams, including SCADA data of the gates opening, rainfall and flow monitoring data from meters in the system, and model results to calculate CSO capture estimates. The dashboard, displayed in Figure 6, allows users to perform wet-weather event reconnaissance and review. Since the pilot study's implementation in June 2021, an estimated 47 MG of CSO have been captured, or an average of 20% CSO reduction per event at Rock Creek, as result of the modified control logic. The Pilot dashboard is a precursor to the RT-DSS and provides MSD a prototype to test functionality and use cases. Currently the dashboard is updated once a month based on static exports of rainfall monitoring and system operational data. Static exports allowed the rapid development of the Pilot Dashboard while data clearance issues are still being worked through. As live connections to data sources are established, the Pilot Dashboard will provide near real-time insights.
Next Steps
The on-going Pilot Study and Existing System Optimization has successfully demonstrated the potential for CSO reduction with modified operations of MSD's existing system. Additional operational strategies recommended by the optimization will be added into the next phase of the Pilot Study and the optimization analysis will expand to two future system scenarios (Interim and Full Build-Out) as defined in Figure 7. With each progressive step, MSD builds consensus towards their RT-DSS vision.