Advanced volt/var management in distribution systems utilizing load forecasting

The increasing penetration of distributed energy resources (DERs), electric vehicles, and variable loads has significantly transformed the operational landscape of modern electrical distribution systems. Traditional Volt/VAr control (VVC) methods, which rely on static and rule based mechanisms, are inadequate in coping with the dynamic and decentralized nature of these networks. This thesis presents an advanced Volt/VAr Management System (VVMS) that integrates short-term load forecasting with intelligent control algorithms to enhance voltage regulation and reactive power balance in distribution systems. The proposed methodology leverages profile-based historical data and predictive modeling to estimate future reactive power requirements and determine optimal switching operations for grid components such as capacitor banks, reactors, and magnetic controlled reactors (MCRs). A modular algorithm structure is developed to calculate power factor deviations, determine compensation type and quantity, and prioritize control devices based on a multi-criteria decision mechanism. By using a tolerance-based control logic and incorporating forecasted load behavior, the system minimizes unnecessary switching, enhances operational reliability, and prolongs equipment life. The methodology is validated using IEEE 33 Bus Test Feeder system, demonstrating significant improvements in voltage profile stability, power loss reduction, and reactive power efficiency. The findings confirm that integrating forecasting with Volt/VAr management provides a scalable, proactive, and cost-effective solution for modern distribution networks facing increasing uncertainty and variability.