Using data-driven approaches to reduce storm sewer overflow spills – CENTAUR use case

<p dir="ltr">Storm Sewer Overflows (SSOs) are a significant public concern, spilling wastewater into rivers and coastal waters during heavy rainfall. There are over 15,000 operational SSOs across England and Wales. The 2021 Environment Act led to Defra’s 2022 Storm Overflows Discharge Reduction Plan (SODRP) [1], setting strict targets to limit spills and to protect people and the environment. Water companies must deliver the largest infrastructure programme in history, requiring £56 billion in investment to meet these targets. The key investment driver for storm overflows is compliance with ‘unusually heavy rainfall events,’ limiting spills to 10 per year on average. Additionally, high-priority sites, such as environmentally sensitive and bathing waters, face stricter targets to enhance protection. This regulatory push is to ensure reduced pollution risk from SSOs across the UK. This paper will describe and present results for CENTAUR, a local autonomous spill reduction system that utilises existing network storage capacity to attenuate flow peaks. The storage capacity is mobilised through active flow control automatically regulated by an AI real-time control system using data from local water level monitoring. CENTAUR uses existing storage capacity and dynamically modifies flow to store and release water in controlled and safe way [2,3,4] during rain events. Developed over eight years, CENTAUR successfully transitioned from innovation to market ready product with water company support.</p><p dir="ltr">This paper was presented at the 21st Computing and Control in the Water Industry Conference (CCWI 2025) at the University of Sheffield (1st - 3rd September 2025).</p>