A novel cyber-physical resilience-based strategy for water quality sensor placement in water distribution networks

Water distribution networks (WDNs) employ contamination warning systems (CWS) with water quality sensors to detect water contamination. If some of these sensors are compromised during a contamination event, the consequences regarding public health can be devastating. With the emerging threat of cyber-attacks on critical infrastructure, WDNs are at risk of cyber-physical attacks that can deliberately contaminate water and concurrently mask the event by cyber-attacking sensors in the CWS. We develop a novel optimization strategy for the placement of water quality sensors that maximizes resilience under cyber-physical attack scenarios, suitable for designing a new sensor layout or upgrading an existing one. The resilient strategy is tested on multiple WDNs, focused on the upgrade problem. Results show improvement in the expected CWS detection performance under cyber-physical attacks, while the performance trade-off under nominal operation is minimal compared to classic strategies, hinting at a good design solution for cyber-wise water utilities.

Automated summary

Water distribution networks use water quality sensors in contamination warning systems, but compromising these sensors during a contamination event can have devastating consequences for public health. There is an emerging threat of cyber-attacks on these networks, which can intentionally contaminate water and mask the event by attacking the sensors. In this study, we develop an optimization strategy for sensor placement that maximizes resilience under cyber-physical attack scenarios, suitable for designing new layouts or upgrading existing ones. The strategy is tested on multiple networks, with results showing improved detection performance under cyber-physical attacks and minimal performance trade-off under nominal operation, suggesting a good design solution for cyber-wise water utilities.