Efforts aimed at the energy transition include the creation of a single integrated energy system through the efficient and coordinated management of different energy carriers.

In this context, the application of ontologies, semantic models, and knowledge graphs offers interesting opportunities for the joint use of information related to different energy sectors or from different information sectors.

The report examines the state of the art of use of the international IEC CIM standard for semantic modeling of energy infrastructure data. The evolution of the current model and incompatibilities with previous versions require an update of software solutions based on this standard implemented by RSE in the past. In addition, focusing on how the model is maintained and developed allows for easier and more consistent implementation
of possible extensions of the model to describe new application domains.

In the perspective of managing a multi-energy network using a single information model, the possibility is introduced of describing the gas network using a model derived from IEC CIM.

Next, the issue of energy infrastructure asset management is addressed, with a focus on the information aspects of this process. The IEC CIM standard includes only a limited number of specific classes for the planning, maintenance and management of network assets until their decommissioning; in contrast, the IFC standard is specifically designed for asset management using Building Information Modeling (BIM). One possible solution is to consider IEC CIM as the central model for creating a single network knowledge graph and, for asset management, establish relations to external classes such as those defined by IFC. For example, this method has been applied for the description of the high-to-medium voltage transformer of primary substations.

In the context of multi-energy systems, the examination and study of energy networks through graph analysis, graph embedding and graph machine learning techniques, based on the representation of energy networks as graphs, is also considered. A preliminary activity is described for analyzing the state of the art of these techniques and preparing training data aimed at developing a graph machine learning algorithm for estimating voltages at the nodes of a power grid without the use of power flow analytical techniques. Finally, processes are considered that involve multienergy spatial planning to support energy system development decisions.

Critical issues were identified in terms of usability and interoperability of the planning tools currently employed. With a view to improving and making these tools more efficient, the information and processing flows used in multienergy planning models for the study and proposal of digitization methods were investigated and described.