The Great Lakes Water Authority (GLWA) is a regional water and sewer authority that services nearly 40 percent of the water customers in Michigan, including the City of Detroit and its surrounding suburbs (approximately 3.2 million customers).

Detroit, like much of the United States, is served by Combined Sewer systems – rainwater runoff and domestic sewage use the same pipes. Due to many stressors, such as aging infrastructure, changing populations, and rapid development in the service area, the sewer and stormwater conveyance system is strained well beyond its design. In general, all the contents of these pipes are treated at a water treatment facility before being discharged. However, in cases of inclement weather such as rain storms, the sewers may run out of space for all the rainwater runoff and the regular sewage. As a result, the system has to discharge its excess, in this case into the Detroit River. When untreated sewage is discharged in this way, it is called a combined sewer overflow. Conventional wisdom is that to avoid combined sewer overflows, sewers must increase their capacity – in other words, MORE storage is needed for rainfall and sewage to occupy until it can be treated at a water treatment facility.

To combat these persistent untreated outflows, and with a vision to create more controlled inflows into the wastewater treatment facility, GLWA launched a project in late 2017 with the University of Michigan’s Real-Time Water Systems Lab lead by Branko Kerkez, Assistant Professor at the University of Michigan’s College of Engineering.  The project was to investigate the application of real-time sensing and dynamic control on existing infrastructure for a much more cost effective problem solution.

The desired outcomes of dynamic, real-time control are to:

• Maximize current storage utilization by dynamically controlling pumps, valves, and gates
• Reduce Combined Sewer Overflows
• Equalize flows to the Wastewater Treatment Plant

Since construction of new Combined Sewer Outfall (CSO) basins (grey infrastructure) and green infrastructure (GI) installations may cost over a billion dollars, real-time control presents a great opportunity to stretch the performance of existing assets.

Using a series of sensors inside the sewer system that can track the sewer’s capacity at different locations, the research team demonstrated that GLWA’s current system can be better utilized to reduce the frequency and volume of Combined Sewer Overflows. In other words, they found that Combined Sewer Overflows have been occurring when additional sewer capacity is available. Rather than increasing the amount of storage with costly infrastructure improvements, the research team was able to demonstrate their ability to use a real-time control approach to significantly improve the existing GLWA wastewater and CSO management system by reducing both the occurrence of CSOs and peak flows going to the treatment facility. These benefits can be achieved without new construction, rather by relying entirely on existing infrastructure, which promises to free up significant capital savings for future investments. The team demonstrated that significant CSO reductions can be achieved – as much 100 million gallons per storm event – by using a cloud-hosted, market-based control algorithm that requires only measurements of water levels and flows at few locations in the sewer system. Unlike other control approaches, U-M’s approach can be implemented with a minimal number of sensors and can be modeled using existing stormwater (SWMM) models. This makes it possible for just about any community to adopt real-time control without expert knowledge. To communicate control recommendations from the controller to the stormwater operator, the team designed a decision support dashboard. The dashboard is web-based and not only gives real-time readouts of measurements from across the stormwater system, it also gives control recommendations to the user as determined by the market-based control algorithm. All algorithms, cloud infrastructure, and control models will be shared open source and available to the public free of cost.

This project was recognized during the Water Environment Federation national conference in 2018 (WEFTEC 2018) held in New Orleans. The team from Great Lakes Water Authority and the University of Michigan won first prize in the new “Leaders Innovation Forum for Technology (LIFT) Intelligent Water Systems Challenge”. The project earned the team a $25,000 prize from Xylem Inc. (Rye Brook, N.Y.). The proposed solution would enable Detroit’s current infrastructure to handle an estimated additional 378.5 million L (100 million gal) with projected savings of about $500 million.