pilots-and-use-cases-overview

I-NERGY Pilots' and Use Cases' overview

I-NERGY Pilots are organized in categories in terms of energy services: Energy Commodities Networks or I-NET (AI for networks optimised operations), Distributed Energy Resources or I-DER (AI for RES generation, buildings, districts and communities), and Energy Efficiency and Non-energy related Services or I-ENEF (AI enabling synergies/ implications on other energy and non-energy domains).

Apart from the category criteria of the pilots, the multi-scale and multi-stakeholder approach will enable the validation and testing in a wide range of contexts, facilitating the development of further AI energy analytics services for any EPES stakeholder involved in the energy value chain.

A brief description of the pilots can be found below:

Pilot 1 [R&D Nester] (Portugal). Exploitation of data related to a total of 70 substations EHV/HV from the Portuguese TSO.
  • Use Case 1: AI for enhanced network assets predictive maintenance, integrating off-grid data with condition-based monitoring. Aimed at applying data analytics techniques to incident report data from transmission system assets (e.g. Operation of Circuit Breakers). These will use extracted insights and models to build optimal condition and predictive based maintenance plans for these assets.
  • Use Case 2: AI for network loads and demand forecasting towards efficient operational planning. Aimed at applying AI forecasting techniques to support internal processes of TSOs, including being applicable in the context of operational and marketing planning. This includes aggregated and disaggregated forecasting of energy load at substation level.

 

Pilot 2 [VEOLIA] (Spain). Data stored in a Hubgrade, storing around 600 GB-recordings per year. Through remote control, it has approximately 3,300 meters distributed throughout the Spanish territory.
  • Use Case 3: AI for energy demand prediction to optimise District Heating Network (DHN) operation. It aims to optimise the operation of the DHN to decrease the energy consumption matching the demand to the production considering weather conditions. Thus, AI will be applied to obtain accurate demand predictions and to improve the network with an optimal O&M reducing the failures and increasing the efficiency.
  • Use Case 4: AI for energy saving verification service, increasing the trust on Energy Performance Contracts. It is aimed to develop accurate predictions of energy and economic savings, while reducing the environmental impact. This will be applied through the Energy Performance Contracts, which are instruments that provide funding of energy efficiency interventions.
  • Use Case 5: AI for multi-energy systems decision-support – Reina Sofía. It aims to optimise the production mix of a multi-energy system considering accurate demand prediction and the weather forecast, to find the best production mix for each moment to maximise the use of renewable energy sources.

 

Pilot 3 [ASM] (Italy). ASM is the distribution system operator of the municipality of Terni, which supplies more than 400 GWh per year to about 65,000 end-users.
  • Use Case 6: Cross-functional AI-based predictive analytics to support integrated DSOs asset management and network operation. It aims to perform predictive maintenance of the distribution grid and make an accurate model of the network in real time.
  • Use Case 7: AI-based consumption and flexibility prediction for local community optimal aggregation and flexibility trading. It is intended to leverage the sources of flexibility in the area to reduce reverse power flow in the electric distribution grid. Thus, the existing flexibility in the system will be estimated, and a marketplace will be created; and consequently, the reverse power flow of the network will be reduced.
  • Use Case 8: AI-based energy-driven and non-energy services. It aims at testing demand response within the distribution network, optimising electrical and non-electrical consumption according to the needs of the network, through optimising the operation and providing flexibility to the network.

 

Pilot 4 [BFP] (Italy). It presents RES generation plants in Plugia region
  • Use Case 9: AI-based IoT-enabled PV module-level portfolio optimal predictive maintenance and PV-enhanced industrial plant optimal operation. It aims to improve the asset management strategy by renovation of long-time consolidated O&M procedures by the combination of the downtime reduction, which comes from the application of new predictive maintenance procedures and improved auto-consumption strategies and a monetary gain rise.

 

Pilot 5 [HERON/ PARITY] (Greece). Focused on multiple home or businesses customers that are equipped with EV charging stations.
  • Use Case 10: AI in EV charging infrastructure. It intends to identify load patterns regarding the real-time distribution and attributes of EV charging transactions in public charging stations. Consumption patterns for the charging transactions will be identified, providing incentives for drivers to distribute charging load evenly across public charging stations and time intervals, monitor EV charging measurements and EV charging occupancy, and proposing pricing changes to the station owner in hourly intervals.

 

Pilot 6 [ZEZ] (Croatia). It consists of a virtual energy community in Croatia, characterised by 100 prosumers with solar PV system in place.
  • Use Case 11: AI for peer-to-peer renewable energy trading in virtual energy community. It is aimed to explore possibilities for peer-to-peer energy trading within the community, and education of citizens and creating favourable social momentum.

 

Pilot 7 [SONCE] (Slovenia). Characterised by a virtual energy community of 220 residents, PV power plant and heat storage.
  • Use Case 12: AI for the Ambient Assisted Living and personal safety / security at home. It aims at conceptualising the non-energy services for personal safety/security and AAL based on the deployed sensor infrastructure and aggregated data from elderly care houses. This will be done by improving the non-energy service (by using fingerprinting methods), improving flexibility (by creating a virtual community with dynamic load management), and cost-optimization.

 

Pilot 8 [REA] (Latvia). Cross-domain integration of a variety of heterogeneous historical and live data on financial performance, underlying energy efficiency impact of the investments, through historical extensive smart meters data integration.
  • Use Case 13: AI for energy efficiency investments de-risking. The aim is to collect and process data from smart meters, and apply Machine Learning algorithms, in order to better predict energy consumption and calculate and monitor the energy savings achieved.

 

Pilot 9 [FAEN] (Spain). Data from Energy Performance Certificates (EPCs) and data from weather stations, complemented with other data such as Copernicus data, cadastral and weather data.
  • Use Case 14: AI for improved Energy Performance Certificates reliability. It aims to integrate vast amounts of data in a harmonized way by means of scalable AI mechanisms to characterise the building stock and its energy consumption. This will make EPCs more reliable and better quality, and will support policy makers in the planning of energy retrofitting planning dates, as well as to support eh EPCs quality checking mechanisms.
  • Use Case 15: AI for predicting the climate change impact in RES and energy demand at regional (local) level. It is aimed to support policy makers in the generation of energy strategies at regional level, in the planning of deployment of renewable sources in an effective manner, contemplating Climate Change impacts, through the long-term prediction of renewable sources potential.