Projects

The first work package will develop static and dynamic simulation models of the power system, which are needed to provide information on stability in each operating state. The static power system model will serve as an interface between ELES' historical measurements and the power system's Dynamic Stability Assessment (DSA) tool. The dynamic power system model will be an extension of the static one and will serve for a detailed analysis of the dynamic stability. The RMS representation of the dynamic conditions will be used. Adequate modelling of all electricity generating units with associated control is crucial in this context. As the design of the project is strongly based on how accurately the power system dynamic model reflects the actual operating conditions, the dynamic model will be properly calibrated and verified by PMU measurements captured by ELES.

In the second work package, we will apply our knowledge and understanding of power system operation to identify key characteristics that can adequately describe the static operating state of a power system. We will further derive indicators that best describe the different types of power system stability. In doing so, we will focus on all the main types of dynamic stability of the EPS, including angular stability under large and small disturbances, voltage stability and frequency stability. Each of these types will be considered separately, taking into account the relevant physical laws. We will establish the links between the characteristics of the steady states with the different types of stability, allowing unambiguous detection and classification of each type of stability.

The central database will contain a number of entries, each entry including stability information (expressed as stability indices) and stationary operating condition data (expressed as characteristics). To support the successful Dynamic Stability Assessment (DSA) of the power system, artificial intelligence techniques will optimise the process of creating the database. This includes the definition of AI functionalities and specifications in close cooperation with power system experts. The development, adaptation and parameterisation of AI methods requires the preparation and quality assurance of data from a central database. Explainable AI concepts will provide transparency and insight into AI decisions. Rigorous laboratory testing will reduce bias and data loss and bridge the gap between the laboratory and the real-world environment.

A central database containing a large number of optimised entries with information on the characteristics of the power system and related data on its potential instabilities forms the backbone of the whole concept. In cooperation with ELES, we will first analyse historical snapshots of operating conditions. It is expected that operating patterns are to some extent repeated. It will be important to identify and harmonise appropriate formats for data exchange between the DSA tool and the existing systems in operation at ELES. We will develop an algorithm for automated calibration of simulation models that will allow the transfer of data from historical ELES snapshots into the system simulation models. This will allow the models to be properly calibrated to reflect the conditions in the actual network. We will perform dynamic simulations of the EPS, apply AI algorithms to optimise the structure of the central database and finally create the database. A large number of simulations will be needed to achieve a sufficient size of the central database, which has a direct impact on the running time of these simulations. From a time, technical and financial point of view, the use of cloud services often seems to be the optimal solution. Therefore, if it proves to be useful/reasonable for the project, we will use a cloud service to achieve an adequate size of the central database within a reasonable timeframe.

We will develop an artificial intelligence technique to quickly scan, find and identify similarities between the actual operating status of the power system and that in the central database. The focus will be on speed of operation, as we want to create a tool that will immediately notify the grid operator of any instability in the power system. The coordination of AI and power systems experts is crucial to define all the specifications and functionalities that the AI method must meet, in particular the speed of implementation. The preparation of input data for the AI models is also important as it will allow for better response times and real-time updating of the database.

The ELES Diagnostic Analysis Centre (DAC) acts as the central infrastructure for data concentration at ELES, equipped with all the information and communication technology needed to run the various algorithms and to ensure the success of the project. The project partners will familiarise themselves with the specificities of the existing hardware and software in the DAC, adapted to the needs of ELES. On this basis, the adaptation of the interactive modules will be carried out. This will be followed by the adaptation and installation of the developed tool for the ongoing assessment of the stability of the power system in the DAC environment. This step requires close collaboration with ELES engineers to ensure consistency with the DAC information and communication infrastructure. In the following, the DAC staff will be familiarised with the on-line stability assessment tool, its functions, operation and algorithms following the basic concept. The operation of the developed tool will be demonstrated and support will be provided by the project implementers in case of any problems.

Work Package 7 will analyse the performance of the Power System Stability Assessment Tool. An in-depth and comprehensive assessment requires the participation of all project partners. In this context, a performance study will be carried out, covering in particular the part of the online stability assessment tool that falls under Phase 2 of the project. We will examine the effectiveness of identifying the operational states of the system both in terms of sufficient similarity and speed. Feedback from DAC staff will provide insight into the adequacy of the implemented algorithms, their transparency and the appropriate notification to the end-user of any instability. The task related to the real testing environment will mainly address the adoption of the DSA tool by ELES and its testing.

The dissemination of project results is crucial for the effective promotion of the activities, but is also closely linked to communication activities. The project consortium is committed to implementing an intensive but clear strategy and to carrying out effective dissemination and communication activities. ELES experts will continuously monitor and comment on the progress of the project, as ELES is considered as the end user of the tool. After the completion of WP7, it is expected that ELES experts will gradually gain experience with the tool for the ongoing stability assessment and will be invited to make suggestions and recommendations for modification and/or further development. Descriptions of the activities and the results obtained will be published on the project website. During the project, the results will also be disseminated through international and national conferences (6 conference papers expected), publications in scientific journals with impact factor (4 papers expected) and social media (LinkedIn, etc.).

Project management and consortium governance will include technical, contractual and financial issues. The governance structure of the consortium will be formally established at the project kick-off meeting. In addition, a kick-off meeting will be organised during the first month of the project and thereafter regular meetings of the project partners will be held at agreed intervals to be determined during the kick-off meeting.