Metrology for reliable power grid data analytics

To ensure Europe’s Green Deal’s aim of reducing greenhouse gas emissions to zero by 2050 requires increasing the amount of electricity generated from renewable energy. Unlike conventional power plants, electricity from solar, water or wind is widely geographically distributed and supplies direct current (DC), requiring converters to match the alternating current (AC) in the grid. These factors, along with demand from new technologies such as electric vehicles, is putting an ever-increasing burden on electricity networks.

To achieve policy targets an EU action plan has stated that digitising the energy system is vital.  Digitisation would provide real-time data on the dynamic behaviour of the electrical system allowing a rapid response to detect abnormal phenomena or instrument faults and prevent instabilities that could lead to blackouts. However, large-scale integration of advanced data-driven methods into networks is hindered by a lack of trust by grid operators in their reliability, and requires a demonstratable, standardised framework for performance assessment of power grid data analytics applications.for performance assessment of power grid data analytics applications.

This project will quantify data quality from power grid sensor networks and develop models and a framework to test and validate candidate data analysis methods, providing at least 15 state-of-the-art methods covering at least five different grid phenomena.  Well-documented public test datasets will be created using real and simulated grid measurements which will be made available to measurement supply chains, standards developing organisations, and end users across Europe. In the long-term, this will increase trust in AI and software monitoring solutions, leading to a more stable, energy efficient grid capable of integrating essential sources of renewable energy.