Breaking the Flow and the Bank: Stealthy Cyberattacks on Water Network Hydraulics

TL;DR

This work analyzes stealthy false data injection attacks on water distribution networks by framing sensor adversaries that must satisfy mass/energy balance, state-estimation convergence, and intrusion-detection bypass. It introduces a full-stealth FDI (FS-FDI) formulation and several constraint-relaxed variants (HU-FDI, HA-FDI, R-FDI), leveraging a weighted least squares state estimator and CUSUM/Chi-squared detectors. Through Net1 and Net3 case studies, it demonstrates that carefully coordinated, locally localized attacks can inflate operational costs (e.g., up to ~$62.74\%$ in Net3) while evading detection and maintaining hydraulic plausibility. The findings underscore a practical vulnerability of WDN monitoring and motivate multi-layer defense strategies combining physical validation with statistical monitoring to harden critical water infrastructure against tailored cyber-physical threats.

Abstract

As water distribution networks (WDNs) become increasingly connected with digital infrastructures, they face greater exposure to cyberattacks that threaten their operational integrity. Stealthy False Data Injection Attacks (SFDIAs) are particularly concerning, as they manipulate sensor data to compromise system operations while avoiding detection. While existing studies have focused on either detection methods or specific attack formulations, the relationship between attack sophistication, system knowledge requirements, and achievable impact remains unexplored. This paper presents a systematic analysis of sensor attacks against WDNs, investigating different combinations of physical constraints, state monitoring requirements, and intrusion detection evasion conditions. We propose several attack formulations that range from tailored strategies satisfying both physical and detection constraints to simpler measurement manipulations. The proposed attacks are simple and local -- requiring knowledge only of targeted sensors and their hydraulic connections -- making them scalable and practical. Through case studies on Net1 and Net3 benchmark networks, we demonstrate how these attacks can persistently increase operational costs and alter water flows while remaining undetected by monitoring systems for extended periods. The analysis provides utilities with insights for vulnerability assessment and motivates the development of protection strategies that combine physical and statistical security mechanisms.

Figures (12)

• Figure 1: Water distribution network architecture illustrating system components, monitoring systems, and security mechanisms. Red arrows indicate attack vectors, demonstrating how different SFDIA variants (FS-FDI: Full-Stealth FDI, HU-FDI: Hydraulics-Unaware FDI, HA-FDI: Hydraulics-Aware FDI, R-FDI: Random FDI) interact with the system's security mechanisms.
• Figure 2: Decision-making flowchart for selecting FDI attack strategies based on available knowledge
• Figure 3: Studied Water Distribution Networks: (a) Net3, (b) Net1
• Figure 4: Full-Stealth FDI attack demonstrating stealthy manipulation of pump operations while maintaining detection avoidance
• Figure 5: Hydraulics-Aware FDI impacts on pump operations and associated costs