The Grid-Predict research project was carried out between 1st December 2015 and 31st December 2018. Besides the Center for Solar Energy and Hydrogen Research Baden-Württemberg (ZSW), the Steinbeis Transfer Centre of Ulm University of Technology - Decentralized Regenerative Energy Systems (STZ) was involved. The associated partner Stadtwerke Ulm/Neu-Ulm Netze GmbH provided the link to an existing section of the distribution network. The expertise and services of ZENNER Hessware made it possible to use smart-meter-compliant transmission paths in the project.
The Grid-Predict project addresses distribution grids that reach their design limits by a high feed-in of decentralized solar power and whose current load condition is only partly known due to missing monitoring.
In the project, methods were developed to correctly determine the grid condition even at high levels of fluctuating feed-in at the low-voltage level. The methods also serve to avoid an otherwise necessary curtailment of power generation output by a suitable control of distributed storage systems. Therefore, advanced prediction algorithms and state estimation methods were developed, implemented and tested.
Among other things, the forecast for solar power generation and household load was revised with a network-based grouping of households. This will reduce insecurities in the load and generation forecasts as well as the missing measurement equipment by aggregating multiple supply connections and statistically estimating the network Status.
Furthermore, methods to ensure the compliance with voltage and power limits by appropriate control of distributed storage systems and generation were developed. These build upon a computation-optimized probabilistic load flow method which was also developed within the project.
The methods for predictive network calculation developed in the project were connected to a control room at the Ulm University of Applied Sciences. The same control room was also connected to simulated households within the lab of ZSW in Stuttgart to demonstrate the usage of Smart-Meter compliant transmission paths.
Finally, the algorithms were tested within the above mentioned framework to show the functionality of the entire systems including all the aspects and developed methods of the project.
The final report lists a number of publications produced within the framework of the project.