Methods for evaluating the implications of the transition to carbon neutrality on the power grid

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:Provides novel model reduction and synthetic grid generation techniques that advance power system modeling for high-renewable scenarios.

🏢実務担当者:Offers grid operators methods to assess reinforcement needs and the effectiveness of coordinated control strategies without costly detailed modeling.

🏛政策担当者:Highlights the need for grid planning reforms and the value of synthetic models in regulatory approval processes for renewable connections.

Today’s power grid infrastructure was not built for a decarbonized energy economy and is consequently slowing the transition to CO2-neutrality. Worldwide, renewable energy projects totaling roughly 3 TW of capacity are delayed for lack of grid-connection capacity; some are never realized. In Germany, 10.5 TWh of renewable generation was curtailed in 2023 due to grid bottlenecks; moreover, since 2012, control settings on more than 300,000 inverters have been adjusted because they posed significant stability risks. Current methods are insufficient to perform the modeling required to derive grid expansion needs and grid-connection requirements at early planning stages. The objective of this thesis is to advance power system modeling to enable the early identification and mitigation of grid-related constraints. The dissertation addresses this challenge with the following research questions from four perspectives. (I) From the perspective of energy system modeling: “How can the complexity of transmission system models be reduced while preserving the original model quality as effectively as possible?” (II) From a transmission system operator perspective: “Under which combinations of short circuit power and penetration of grid-forming converters do state-of-the-art distribution system reduction methods fail to model instabilities?” (III) From the private sector and policymaking points of view: "Do synthetically generated grid models allow realistic estimation of distribution grid reinforcement needs?” (IV) From the perspective of distribution system operators: “To what extent do cross-voltage-level operation control strategies reduce the need for distribution grid reinforcement?” By comprehensively addressing gaps in power grid modeling, grid-related barriers to the transition can be avoided. For (I), a reduction methodology is developed that simplifies transmission grid models based on the electrical distance between grid nodes. Model fidelity is evaluated using defined metrics and compared with an established method from the literature. For (II), a dynamic distribution grid model is built; in a sensitivity analysis over the short circuit level provided by the transmission grid and the penetration of grid-forming inverters, system stability modeling of the original network is compared against three established reduction methods and one newly developed method. Quality is assessed using metrics from literature. In (III), expansion needs for four subgrids from four distribution system operators are determined both with the operator’s model and with two synthetic low-voltage grid models generated using different methods; repeated probabilistic disaggregation of the scenarios is used to statistically characterize the estimation error. For (IV), a medium-voltage grid with underlying low-voltage grids is modeled. Scenarios for renewable penetration are created, probabilistically disaggregated, and power flow calculations are performed for four control strategies — each with and without on-load tap-changing distribution transformers; local and communication-based approaches are contrasted. The results show: First, the reduction methodology based on electrical distance outperforms the comparator in most defined quality metrics and captures interactions between energy system development and the transmission grid more accurately at the same model complexity (I). Second, the distribution grid reduction method most widely used by transmission system operators sometimes fails to detect instabilities; all three established procedures violate existing quality criteria. The newly developed reduction method meets these criteria (II). Third, synthetic low-voltage grid models enable robust attribution of overloads to additional loads or generators; one of the methodologies reliably classifies the magnitude of distribution grid expansion needs (III). Fourth, communication-based, cross-voltage-level control strategies can increase the hosting capacity of medium- and low-voltage grids and thereby connect additional loads and generators without grid reinforcement (IV). The work thus makes four contributions: it provides scalable model reduction techniques, identifies the limits of common reductions in stability assessment, and demonstrates the value of synthetic grids as well as coordinated control for planning and operational decisions. In practice, the results can help grid operators and policymakers identify reinforcement needs early, tailor connection requirements, and reduce rejections of grid connection requests.

• openalex https://doi.org/10.6094/unifr/283607first seen 2026-06-14 04:43:08 · last seen 2026-06-16 04:50:12