Practical example and current challenges
- Cheap wind power from the north does not reach consumers in the south due to a bottleneck in the power grid. This case leads to the wind turbine going off the grid and is called “redispatch”. "Redispatch" refers to the change in the planned use of power plants to avoid grid bottlenecks. Individual solutions and platforms are currently being set up for the regulatory Redispatch 2.0, which will be mandatory from October 2021.
- Redispatch 2.0 is currently converting the grid management processes to a planned value-based procedure. The reaction to acute grid bottlenecks should no longer be a spontaneous downgrade of wind and PV parks. Instead, a forecast-based grid bottleneck avoidance process between grid operators and plant deployment managers should be used. This primarily affects large systems.
- This is only a first step towards the decentralized and AI-based grid bottleneck reduction procedure Redispatch 3.0, in which small private systems are also to be integrated into the forecast and grid optimization processes. In this context, a larger and more integrated database is required. Relevant data form the basis for AI algorithms and the integration of a new SMGW infrastructure (Smart Meter Gateway) for BSI-compliant, secure IoT communication with potentially controllable energy generation and consumption units.
- Functionally, this involves the migration of Redispatch 2.0 modules to federated cloud infrastructures, increased data sharing between grid operators and other actors, as well as the transformation of classic SCADA-based (Supervisory Control and Data Acquisition) remote control technologies for grid control to innovative IoT infrastructures, which will be secured by SMGW-infrastructure.
- In the future, in addition to classic grid control systems (NLS), a cloud-based and BSI-compliant IoT alternative for direct communication with decentralized generation and consumption units with a nominal output of less than 100 kW must be set up as part of Redispatch 3.0. In most cases, these small systems are currently not accessible in terms of communication and will only be accessible as communication endpoints via CLS functionalities of the intelligent metering systems (iMSys) in the course of the upcoming smart metering rollout.
What added value does the "GAIA-X project" offer?
- In addition to a scalable and secure IoT alternative, the GAIA-X infrastructure enables a significantly wider use of AI / machine learning (ML) technologies to generate different types of forecast, determine Redispatch’s potentials and implement control interventions in several thousand decentralized producers and consumers.
- In addition, edge computing and cloud meshes are to be used to distribute the system-critical functions to the GAIA-X nodes and edge computing units of the grid operators in such a way that power-related unavailability of GAIA-X nodes in one region do not affect the proper Redispatch-operation in other regions.
- The implementation of Redispatch processes in the energy sector enables an introduction and significant improvement of the AI / ML-based forecasting and provisioning planning algorithms in redispatch processes. Thus, the grid utilization can be optimized. In addition, experience is gained using edge compute capabilities and the federated approach of the GAIA-X nodes for redispatch solutions to prepare a migration of further OT-oriented use cases of the energy domain to cloud technologies. Specific other applications are in particular: Decentralized training of AI / ML algorithms for forecasts, condition monitoring, more scalable and highly automated control systems from the cloud, as well as the creation of new energy – flexibility – products and markets for system operators.
Use Case Team
- Sebastian Lehnhoff – OFFIS
- Johannes Dorfner – OFFIS