CSO Reduction by Integrated Model Predictive Control of Stormwater Inflows: A Simulated Proof of Concept Using Linear Surrogate Models

Abstract

Combined sewer overflows (CSO) of mixed stormwater and wastewater pollute nearby receiving surface waters and pose a risk to the environment and human health. We use “integrated stormwater inflow control” to mitigate CSO by dynamically controlling the inflow of stormwater to the combined sewer system in real time, expanding the physical space of traditional real‐time control. This control is carried out with model predictive control (MPC), which we base on convex optimization including a linear internal surrogate model of the controllable aboveground and belowground infrastructure. A detailed hydrodynamic model is used to evaluate the results and recursively initialize the surrogate model. MPC dynamically decides when to let stormwater enter the sewer system and when to store and convey excess stormwater in the aboveground infrastructure otherwise intended for passive cloudburst management. The performance was quantified in a simulation study in Copenhagen, Denmark, using a 1‐D distributed hydrodynamic model and 32 rain events from 2016, of which 18 caused CSO in the situation without control. Four of the 18 CSO events were avoided with MPC, and the total CSO volume was reduced by 98.4% of the potential reducible volume. For one event, stormwater was unnecessarily kept on the surface because the surrogate model wrongly predicted a CSO. The computational cost was in all cases compatible with an operational implementation. With the invention of proper actuators for control of stormwater inflows, we show that MPC of stormwater inflows may be a viable supplement to more traditional passive ways of managing stormwater in urban areas.

Combined Sewer Overflow and Flooding Mitigation Through a Reliable Real‐Time Control Based on Multi‐Reinforcement Learning and Model Predictive Control

“…The CSO and flooding volume are affected by the diversity of rainfall events, even under the same CSS (Lund et al, 2018(Lund et al, , 2020van Daal et al, 2017). Thus, they might not be objective enough to measure the performance of RLs.…”

“…Model predictive control (MPC), one of the widely studied optimization‐based methods, uses internal UDS model, rainfall forecasts, and optimization algorithm to recursively calculate optimal control action, and thus, achieves better adaptability and performance compared with HC (Fu et al., 2008; Joseph‐Duran et al., 2015; Lund et al., 2020; Schütze et al., 2002; Sun et al., 2020). It consists of receding horizon principle and optimization.…”

Flooding and Overflow Mitigation Using Deep Reinforcement Learning Based on Koopman Operator of Urban Drainage Systems

“…The model predictive control (MPC), a representative method of the optimization‐based approaches, uses an internal UDS model and real‐time rainfall forecasts to recursively compute optimal control variables on receding horizons and, thus, avoids the drawbacks of the heuristic methods. However, it is constrained by the computation time of the real‐time optimization and the accuracy of the rainfall forecast (Lund et al., 2020).…”

The Value of Aggregated City Scale Models to Rapidly Assess SuDS in Combined Sewer Systems

“…The state-of-the-art in research has demonstrated that a reduced complexity approach to CSO modelling can be used to create accurate simulations in a fraction of the time (Nielsen et al, 2007;Morales et al, 2017). This enables modelling of CSOs in contexts that require rapid simulations, for example, Lund et al (2020) use surrogate modelling to control CSO behaviour with optimisation of Model Predictive Control (MPC). These studies all employ some form of spatial aggregation to achieve the reduction in complexity.…”