Artificial Intelligence for Managing Uncertainty in Energy Supply and Demand

This is the challenge faced by one of the top players in the energy sector, with high, medium, and low voltage infrastructure distributed over one of the largest metropolitan areas in Italy. To address this challenge, the company entrusted the consortium led by the Lutech Group with the implementation of an Advanced Smart Grid Management & Metering System.

This new energy distribution management system is capable of centrally representing the entire network's structure and collecting a vast array of information on its status and usage through IoT sensors and Digital Twin simulations. It monitors, detects, and processes the collected data, leveraging innovative Data Science and Artificial Intelligence technologies.

The need was to implement a network digitalization program to improve asset efficiency and reduce interruptions and financial penalties from Arera, aiming to centralize a unified view of the current state and the continuity needs of the energy service.

The ASGMMS (Advanced Smart Grid Management & Metering System), integrated with SCADA (which detects all field parameters, from power to capacity) and GIS (which represents the Client's assets), constitutes a truly "intelligent" system capable of aggregating all relevant information, speeding up fault location and isolation, state estimation, and voltage regulation across the network. This improves resource efficiency and reduces the mean time to repair.

In this context, AI is being experimented with autonomous driving algorithms for asset topology discovery and rapid fault management, with the goal of:

• Improving network observability: Through the unified system and leveraging IoT sensors and the Digital Twin paradigm for optimized network monitoring and control, enabling greater transparency of the behavior of power sources and more effective and faster selection of corrective actions in response to identified issues.
• Increasing network resilience: Thanks to advanced data processing technologies (including AI), critical factors are immediately highlighted, and there is also a preventive signaling of critical network congestion conditions through alarms regulated by preset but variable thresholds. The system allows for the planning and implementation of evolutionary and maintenance interventions to reduce risks.
• Managing network flexibility: By acquiring data and controlling Smart Metering solutions (IoT-based "smart meters"), it is possible to determine and support the balancing of energy flows and enable innovative energy use flexibility services, balancing the meeting of supply and demand.

Predictive analysis, based on AI and Advanced Analytics, along with the collection of service quality indicators, will allow us to anticipate system dynamics: it will be possible to plan and implement not only a "reactive" solution to the current problem but also to ensure the robustness of the network through preventive measures and maintenance planning, including evolutionary structural interventions on the network.

Ensuring maximum resource efficiency without impacting service quality, through algorithms for state estimation and power flow calculation, along with a significant reduction in average repair times, are the success factors of this innovative project aimed at managing the uncertainty of energy supply and demand, a physiological characteristic of a context dominated by the use of renewables and electrification.