The project established a comprehensive business requirements framework, mapping five core objectives to EU policy priorities. For each of the nine GRAVITEQA tools, detailed functional and non-functional requirements were specified per ISO/IEC 25010 with MoSCoW prioritisation. Nine use cases were documented using the IEC 62559 template and positioned within the Smart Grid Architecture Model. Twenty-six Key Performance Indicators were defined with their derivation methodology, alongside a complete data-collection plan. A stakeholder survey identified and classified barriers, producing actionable strategies for market adoption.
Preliminary Life Cycle Assessment and Techno-Economic Analysis showed that reusing infrastructure from decommissioned coal plants for gravitational storage yields substantial environmental benefits, with iron-based configurations emerging as most economically promising. The lifecycle importance of edge-device manufacturing was highlighted, and initial cost assessments for quantum approaches identified scalability limitations. Shore-side electrification analysis demonstrated measurable CO2 savings and clarified conditions affecting financial viability.
Mathematical formulations were developed for Load-side Management for Flexibility Provisioning and Facility Location Allocation, advancing quantum and quantum-inspired computing applications in the energy sector. The first alpha implementations of quantum-inspired solutions and toy-case quantum computing solutions were completed, alongside a solver-agnostic formulation software in alpha version.
A Nodal Flexibility Region Estimation methodology was developed, enabling fast, grid-feasible flexibility procurement from distributed energy resources, with the network model verified against established power-flow tools. An end-to-end trustworthy learning framework for non-convex optimization was initiated, combining machine learning with explicit differentiation to guarantee feasible solutions in real-time, incorporating explainable AI principles via SHAP-based transparency.
A conformal prediction framework for cold-ironing demand forecasting was developed, providing prediction intervals with coverage guarantees for intra-day and day-ahead horizons. A port-asset scheduling algorithm was formulated to minimise energy costs and support demand response while respecting grid constraints. An AIS-data-driven seaport electrification planning methodology was developed and demonstrated using Port of Piraeus data, covering traffic analysis, demand estimation, emissions assessment, cost-benefit evaluation, and grid-impact simulation.
A structured three-stage methodology was developed for transforming decommissioned coal plants and mines into long-duration energy storage facilities. Mathematical formulations were established for shaft-based and piston-based gravitational storage, with a capacity estimation algorithm to determine technically feasible configurations. A financial assessment model incorporating investment costs, revenues from energy arbitrage, and potential hybrid configurations with pumped hydro storage was formulated.
A unified testing and evaluation framework was defined with three testing levels: component verification, tool validation, and use-case validation. A cold-ironing validation dataset was constructed for the Port of Varna using AIS data. A benchmarking strategy was defined comparing Linear Regression against Random Forest, XGBoost, LightGBM, and MLP, with two testing scenarios specified for the 36-hour forecasting horizon. A validation roadmap for circularity and storage technologies was also developed, mapping use cases to technical, economic, grid, and environmental KPIs.
In summary, the first 18 months delivered clear requirements, defined KPIs, alpha versions of all tools, and a complete validation roadmap. The project is now positioned for refinement, integration, and rigorous validation against its ambitious objectives.
