SCADA + SWMM - HYDROLOGY = Data driven modeling of SSO 700 storage and treatment facility in Cincinnati

Reese Johnson, Melissa Gatterdam, Bryant E. McDonnell, Ruben Kertesz, Fred Myers, Luis Montestruque and Elena Rubchinskaya

ABSTRACT

The Metropolitan Sewer District of Greater Cincinnati has been making strides toward reducing overflows using advancements in their smart sewers. Improvements within the Mill Creek drainage network include real time remote monitoring, coordinated control of inline storage facilities, and the full automation of the SSO 700 Storage and Treatment Facility (STF). Opportunities were revealed to further leverage the SSO 700 STF to reduce surcharging and overflow volumes within the downstream interceptor. The facility was adapted with remote reactive control logic to begin diverting flow before the interceptor is hydraulically stressed.

The biggest question that comes up after a new strategy is implemented is “How can we continue to tune this to make it even better?!?! We collect a lot of data! How can we use it more effectively?!?!?” When designing control strategies, normally this is done using a rainfall-driven hydraulic model. After implementing the controls into the real-world collection system, beyond using available data to interpret and study the efficacy of the prescribed control solution, there has not been an accurate way to understand system-scale impacts for real wet weather events. This is because rainfall-driven hydraulic models have significant uncertainty in rainfall inputs and runoff surface hydrology.

The overcome these limitations, a Level2Flow plug-in was created for SWMM5, using the expansive API (OpenWaterAnalytics SWMM / PySWMM). The plug-in solves for network inflows using the sensor levels without the need for rainfall. For example, if the reported level is 3ft, the utility will solve for that level by iteratively stepping up/stepping down the inflow rate at that point in the system. Solved inflow points based on observed level data matched with NSE > 0.95, suggesting an excellent representation. Once the inflows at all the boundary conditions are determined from “today’s” data, they are injected into an alternative controls simulation. The plug-in provides an optimization framework to quantify how the control logic is performing, and how it would have performed for the same wet weather event but different logic, with high confidence.


Permanent link: