The progressing integration of renewables (RE) into distribution grids leads to bidirectional power flows. Due to the fact that the power infeed by RE exceeds the grid capacities, bottlenecks occur increasingly and thereby renewable power generation has to be curtailed. For the decision if countermeasures like flexibility options could be applied, the knowledge of the amount of curtailed power as well as its local and temporal availability is crucial. This study proposes an approach to determine the curtailed power characteristics, which creates time series simulating the potential vertical power flow at medium voltage/ high voltage (MV/HV) transformers without curtailment. The approach is thereby modular and adjustable to the investigated (or any other) grid, so it achieves a high precision. The power flow at these transformers is modelled by load profiles and power curves of connected RE, which are adjusted individually for each transformer on the basis of the corresponding historical power flow. For validation, the transformers are clustered corresponding to their power flow characteristics and the uncertainty is determined for each cluster. Modelled wind power shows a mean deviation below 2% of total installed wind power. Thereby, the presented modelling approach is able to determine the curtailed power in regions with wind power-related curtailments.
Renewable energies curtailment induced by grid congestions increase due to grown renewable energies integration and the resulting mismatch of grid expansion. Short-term predictions for curtailment can help to increase the efficiency of its management. This paper proposes a novel, holistic approach of a short-term curtailment prediction for distribution grids. The load flow calculations for congestion detection are realized by taking different operational security criteria into account, whereas the models for the node-injections are adjusted to the characteristic of each grid node specifically. The determination of required curtailment based on the resulting congestions considers uncertainties of component loading and its corresponding probability. The forecast model is validated using an actual 110 kV distribution grid located in Germany. In order to meet the requirements of a forecast model designed for operational business, prediction accuracy, and its greatest source of error are analyzed. Furthermore, a suitable length of training data is investigated. Results indicate that a six month time period for maintenance gains the highest accuracy. Curtailment prediction accuracy is better for transmission system operator components than for distribution system operator components, but the Sørensen Dice factor for the aggregated grid shows a high match of historic and predicted curtailment with a value of 0.84 and a low error for curtailed energy, which makes 2.23% of the historic curtailed energy. The model is a promising approach, which can contribute to improvement of curtailment strategies and enable valuable insight into distribution grids.
The growing integration of renewable energies into electricity grids leads to an increase of grid congestions. One countermeasure is the curtailment of renewable energies, which has the disadvantage of wasting energy. Forecasting congestion provides valuable information for grid operators to prepare and instruct countermeasures to reduce these energy losses. This paper presents a novel approach for congestion prediction in distribution grids (i.e. up to 110 kV) considering the n-1 security criterion. For this, our method considers node injections and power flow and combines three artificial neural network models. The analysis of study results shows that the implemented neural networks within the presented approach perform better than naive forecasts models. In the case of vertical power flow, the artificial neural networks also show better results than comparable parametric models: average values of the mean absolute errors relative to the parametric models range from 0.89 to 0.21. A high level of accuracy can be achieved for the neural network that predicts the loading of grid components with a F1 score of 0.92. Further, also with a F1 score of 0.92, this model shows higher accuracy for the distribution grid components than for those of the transmission grid, which achieve a F1 score of 0.84. The presented approaches show good potential to support grid operators in congestion management.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.