Dynamic IoT–PID Control for Energy-Efficient Water Distribution: EPANET-Based Digital Twin Validation in Varied Geographical Terrains

Abstract

Topographical heterogeneity in water distribution networks frequently causes pressure imbalance, hydraulic inefficiency, and elevated energy consumption, particularly in regions with significant elevation gradients. This study aims to develop and validate a dynamic Internet of Things (IoT)-based pressure control model within a cyber–physical system framework for energy-efficient water distribution under varied geographical conditions. The primary contribution of this work lies in the separation of strategic and tactical control layers, where a Digital Twin based on EPANET dynamically generates optimal pressure setpoints, while distributed proportional–integral–derivative controllers execute real-time valve regulation at the network edge. The research adopts a Design Science Research methodology to design, implement, and evaluate a four-layer architecture consisting of physical sensing and actuation, long-range communication, tactical control, and strategic simulation layers. Validation is conducted using EPANET-based simulations across three control scenarios: a baseline condition without dynamic control, a static rule-based valve control scenario, and the proposed dynamic IoT–PID control scenario. The experimental procedure involves comparative analysis using control performance metrics including overshoot, settling time, steady-state error, and root mean square error. Simulation results demonstrate that the baseline configuration suffers from severe pressure imbalance and hydraulic backflow, while static rule-based control partially mitigates inefficiencies but fails to adapt to demand variability. In contrast, the proposed dynamic IoT–PID approach achieves precise pressure regulation with overshoot below 2% and tracking error maintained under 0.5 meters across all evaluated scenarios. These findings confirm that integrating a Digital Twin with real-time PID control significantly improves pressure stability and operational efficiency. The proposed architecture offers practical implications for smart water infrastructure in geographically diverse regions, providing a scalable foundation for adaptive pressure management, energy optimization, and future digital-twin-driven water distribution systems.