arXiv subjects

Frank Hellmann

Publications and source records attributed to Frank Hellmann.

At least 37 records · Page 2Linked to original sources

arXiv2024

Complex Couplings -- A universal, adaptive and bilinear formulation of power grid dynamics

The paper is now published in PRX Enegry. Please refer to the PRX version from now on. Anna Büttner and Frank Hellmann. "Complex Couplings-A Universal, Adaptive, and Bilinear Formulation of Power Grid Dynamics." The energy transition introduces new classes of dynamical actors into the power grid. There is a growing need for so-called grid-forming inverters (GFIs) that can contribute to dynamic grid stability as the share of synchronous generators decreases. Understanding the collective behavior and stability of future grids, featuring a heterogeneous mix of dynamics, remains an urgent and challenging task. Two recent advances in describing such modern power grid dynamics have made this problem more tractable: First, the normal form for grid-forming actors provides a uniform, technology-neutral description of plausible grid dynamics, including grid-forming inverters and synchronous machines. Secondly, the notion of the complex frequency has been introduced to effortlessly describe how the nodal dynamics influence the power flows in the grid. The major contribution of this paper is to show how the normal form approach and the complex frequency dynamics of power grids combine, and how they relate naturally to adaptive dynamical networks and control affine systems. Using normal form and complex frequency, we derive a remarkably elementary and universal equation for the collective grid dynamics. Notably, we obtain an elegant equation entirely in terms of a matrix of complex couplings, in which the network topology does not explicitly appear. These complex couplings give rise to new adaptive network formulations of future power grid dynamics. We give a new formulation of the Kuramoto model with inertia as a special case.

nlin.AO↗

arXiv2023

Toward Dynamic Stability Assessment of Power Grid Topologies using Graph Neural Networks

To mitigate climate change, the share of renewable energies in power production needs to be increased. Renewables introduce new challenges to power grids regarding the dynamic stability due to decentralization, reduced inertia, and volatility in production. Since dynamic stability simulations are intractable and exceedingly expensive for large grids, graph neural networks (GNNs) are a promising method to reduce the computational effort of analyzing the dynamic stability of power grids. As a testbed for GNN models, we generate new, large datasets of dynamic stability of synthetic power grids, and provide them as an open-source resource to the research community. We find that GNNs are surprisingly effective at predicting the highly non-linear targets from topological information only. For the first time, performance that is suitable for practical use cases is achieved. Furthermore, we demonstrate the ability of these models to accurately identify particular vulnerable nodes in power grids, so-called troublemakers. Last, we find that GNNs trained on small grids generate accurate predictions on a large synthetic model of the Texan power grid, which illustrates the potential for real-world applications.

cs.LG↗

arXiv2023

Protecting the Texas power grid from tropical cyclones: Increasing resilience by protecting critical lines

The Texan electric network in the Gulf Coast of the United States is frequently hit by Tropical Cyclones (TC) causing widespread power outages, a risk that is expected to substantially increase under global warming. Here, we introduce a new approach of combining a probabilistic line fragility model with a network model of the Texas grid to simulate the temporal evolution of wind-induced failures of transmission lines and the resulting cascading power outages from seven major historical hurricanes. The approach allows reproducing observed supply failures. In addition, compared to a static approach, it provides a significant advantage in identifying critical lines whose failure can trigger large supply shortages. We show that protecting only 1% of total lines can reduce the likelihood of the most destructive type of outages by a factor of between 5 and 20. The proposed modelling approach could represent a tool so far missing to effectively strengthen the power grids against future hurricane risks even under limited knowledge.

physics.soc-ph↗

arXiv2022

Towards dynamic stability analysis of sustainable power grids using graph neural networks

To mitigate climate change, the share of renewable needs to be increased. Renewable energies introduce new challenges to power grids due to decentralization, reduced inertia and volatility in production. The operation of sustainable power grids with a high penetration of renewable energies requires new methods to analyze the dynamic stability. We provide new datasets of dynamic stability of synthetic power grids and find that graph neural networks (GNNs) are surprisingly effective at predicting the highly non-linear target from topological information only. To illustrate the potential to scale to real-sized power grids, we demonstrate the successful prediction on a Texan power grid model.

cs.LG↗

arXiv2022

Linear Response Theory for Renewable Fluctuations in Power Grids with Transmission Losses

We study the spreading of renewable power fluctuations through grids with Ohmic losses on the lines. By formulating a network adapted linear response theory, we find that vulnerability patterns are linked to the left Laplacian eigenvectors of the overdamped eigenmodes. We show that for tree-like networks fluctuations are amplified in the opposite direction of the power flow. This novel mechanism explains vulnerability patterns that were observed in previous numerical simulations of renewable micro-grids. While exact mathematical derivations are only possible for tree like networks with homogeneous response, we show that the mechanisms discovered also explain vulnerability patterns in realistic heterogeneous meshed grids by studying the IEEE RTS-1996 test system.

nlin.AO↗

arXiv2022

Predicting Basin Stability of Power Grids using Graph Neural Networks

The prediction of dynamical stability of power grids becomes more important and challenging with increasing shares of renewable energy sources due to their decentralized structure, reduced inertia and volatility. We investigate the feasibility of applying graph neural networks (GNN) to predict dynamic stability of synchronisation in complex power grids using the single-node basin stability (SNBS) as a measure. To do so, we generate two synthetic datasets for grids with 20 and 100 nodes respectively and estimate SNBS using Monte-Carlo sampling. Those datasets are used to train and evaluate the performance of eight different GNN-models. All models use the full graph without simplifications as input and predict SNBS in a nodal-regression-setup. We show that SNBS can be predicted in general and the performance significantly changes using different GNN-models. Furthermore, we observe interesting transfer capabilities of our approach: GNN-models trained on smaller grids can directly be applied on larger grids without the need of retraining.

physics.soc-ph↗

arXiv2022

Ambient Forcing: Sampling Local Perturbations in Constrained Phase Spaces

Ambient Forcing is a novel method to sample random states from manifolds of differential-algebraic equations (DAE). These states can represent local perturbations of nodes in power systems with loads, which introduces constraints into the system. These states must be valid initial conditions to the DAE, meaning that they fulfill the algebraic equations. Additionally, these states should represent perturbations of individual variables in the power grid, such as a perturbation of the voltage at a load. These initial states enable the calculation of probabilistic stability measures of power systems with loads, which was not yet possible, but is important as these measures have become a crucial tool in studying power systems. To verify that these perturbations are network local, i.e. that the initial perturbation only targets a single node in the power grid, a new measure, the spreadability, related to the closeness centrality, is presented. The spreadability is evaluated for an ensemble of typical power grids. The ensemble depicts a set of future power grids where consumers, as well as producers, are connected to the grid via inverters. For this power grid ensemble, we additionally calculate the basin stability as well as the survivability, two probabilistic measures which provide statements about asymptotic and transient stability. We also revisit topological classes, that have been shown to predict the basin stability of grids and explore if they still hold for grids with constraints and voltage dynamics. We find that the degree of the nodes is a better predictor than the topological classes for our ensemble. Finally, ambient forcing is applied to calculate probabilistic stability measures of the IEEE 96 test case.

nlin.AO↗

arXiv2022

A normal form for grid forming power grid actors

Future power grids will be operating a large number of heterogeneous dynamical actors. Many of these will contribute to the fundamental dynamical stability of the system. By taking a complexity theoretic perspective we derive a normal form for grid forming components in power grids. This allows analyzing the grids systemic properties without the need for detailed technical models. Our approach is based on the physics of the power flow in the grid on the one hand, and on the common symmetry that is inherited from the control objectives grid-forming power grid components are trying to achieve. We provide a first experimental validation that this normal form can capture the behavior of complex grid forming inverters without any knowledge of the underlying technology, and show that it can be used to make technology independent statements on the stability of future grids.

nlin.AO↗

arXiv2021

An Open Source Software Stack for Tuning the Dynamical Behavior of Complex Power Systems

BlockSystems.jl and NetworkDynamics.jl are two novel software packages which facilitate highly efficient transient stability simulations of power networks. Users may specify inputs and power system design in a convenient modular and equation-based manner without compromising on speed or model detail. Written in the high-level, high-performance programming language Julia a rich open-source package ecosystem is available, which provides state-of-the-art solvers and machine learning algorithms. Motivated by the recent interest in the Nordic inertia challenge we have implemented the Nordic5 test case and tuned its control parameters by making use of the machine learning and automatic differentiation capabilities of our software stack.

eess.SY↗

arXiv2021

Probabilistic Behavioral Distance and Tuning - Reducing and aggregating complex systems

Given a complex system with a given interface to the rest of the world, what does it mean for a the system to behave close to a simpler specification describing the behavior at the interface? We give several definitions for useful notions of distances between a complex system and a specification by combining a behavioral and probabilistic perspective. These distances can be used to tune a complex system to a specification. We show that our approach can successfully tune non-linear networked systems to behave like much smaller networks, allowing us to aggregate large sub-networks into one or two effective nodes. Finally, we discuss similarities and differences between our approach and $H_\infty$ model reduction.

math.OC↗

arXiv2021

Moving the epidemic tipping point through topologically targeted social distancing

The epidemic threshold of a social system is the ratio of infection and recovery rate above which a disease spreading in it becomes an epidemic. In the absence of pharmaceutical interventions (i.e. vaccines), the only way to control a given disease is to move this threshold by non-pharmaceutical interventions like social distancing, past the epidemic threshold corresponding to the disease, thereby tipping the system from epidemic into a non-epidemic regime. Modeling the disease as a spreading process on a social graph, social distancing can be modeled by removing some of the graphs links. It has been conjectured that the largest eigenvalue of the adjacency matrix of the resulting graph corresponds to the systems epidemic threshold. Here we use a Markov chain Monte Carlo (MCMC) method to study those link removals that do well at reducing the largest eigenvalue of the adjacency matrix. The MCMC method generates samples from the relative canonical network ensemble with a defined expectation value of $λ_{max}$. We call this the "well-controlling network ensemble" (WCNE) and compare its structure to randomly thinned networks with the same link density. We observe that networks in the WCNE tend to be more homogeneous in the degree distribution and use this insight to define two ad-hoc removal strategies, which also substantially reduce the largest eigenvalue. A targeted removal of 80\% of links can be as effective as a random removal of 90\%, leaving individuals with twice as many contacts.

physics.soc-ph↗

arXiv2021

NetworkDynamics.jl -- Composing and simulating complex networks in Julia

NetworkDynamics.jl is an easy-to-use and computationally efficient package for working with heterogeneous dynamical systems on complex networks, written in Julia, a high-level, high-performance, dynamic programming language. By combining state of the art solver algorithms from DifferentialEquations.jl with efficient data structures, NetworkDynamics.jl achieves top performance while supporting advanced features like events, algebraic constraints, time-delays, noise terms and automatic differentiation.

cs.MS↗

arXiv2021

Probabilistic Stability Assessment for Active Distribution Grids

This paper demonstrates the concept of probabilistic stability assessment on large-signal stability in the use case of short circuits in an active distribution grid. Here, the concept of survivability is applied, which extends classical stability assessments by evaluating the stability and operational limits during transients for a wide range of operating points and failures. For this purpose, a free, open-source, and computationally efficient environment (Julia) for dynamic simulation of power grids is used to demonstrate its capabilities. The model implementation is validated against established commercial software and deviations are minimal with respect to power flow and dynamic simulations.The results of a large-scale survivability analysis reveal i) a broad field of application for probabilistic stability analysis and ii) that new non-intuitive stability correlations can be obtained. Hence,the proposed method shows strong potential to efficiently conduct power system stability analysis in active distribution grids.

eess.SY↗

arXiv2021

Relative Canonical Network Ensembles -- (Mis)characterizing Small-World Networks

What do generic networks that have certain properties look like? We define Relative Canonical Network ensembles as the ensembles that realize a property R while being as indistinguishable as possible from a generic network ensemble. This allows us to study the most generic features of the networks giving rise to the property under investigation. To test the approach we apply it first to the network measure "small-world-ness", thought to characterize small-world networks. We find several phase transitions as we go to less and less generic networks in which cliques and hubs emerge. Such features are not shared by typical small-world networks, showing that high "small-world-ness" does not characterize small-world networks as they are commonly understood. On the other hand we see that for embedded networks, the average shortest path length and total Euclidean link length are better at characterizing small-world networks, with hubs that emerge as a defining feature at low genericity. We expect the overall approach to have wide applicability for understanding network properties of real world interest.

physics.soc-ph↗

arXiv2021

PowerDynamics.jl -- An experimentally validated open-source package for the dynamical analysis of power grids

PowerDynamics.jl is a Julia package for time-domain modeling of power grids that is specifically designed for the stability analysis of systems with high shares of renewable energies. It makes use of Julia's state-of-the-art differential equation solvers and is highly performant even for systems with a large number of components. Further, it is compatible with Julia's machine learning libraries and allows for the utilization of these methods for dynamical optimization and parameter fitting. The package comes with a number of predefined models for synchronous machines, transmission lines and inverter systems. However, the strict open-source approach and a macro-based user-interface also allows for an easy implementation of custom-built models which makes it especially interesting for the design and testing of new control strategies for distributed generation units. This paper presents how the modeling concept, implemented component models and fault scenarios have been experimentally tested against measurements in the microgrid lab of TECNALIA.

eess.SY↗

arXiv2020

Iterative learning control in prosumer-based microgrids with hierarchical control

Power systems are subject to fundamental changes due to the increasing infeed of renewable energy sources. Taking the accompanying decentralization of power generation into account, the concept of prosumer-based microgrids gives the opportunity to rethink structuring and operation of power systems from scratch. In a prosumer-based microgrid, each power grid node can feed energy into the grid and draw energy from the grid. The concept allows for spatial aggregation such that also an interaction between microgrids can be represented as a prosumer-based microgrid. The contribution of this work is threefold: (i) we propose a decentralized hierarchical control approach in a network including different time scales, (ii) we use iterative learning control to compensate periodic demand patterns and save lower layer control energy and (iii) we assure asymptotic stability and monotonic convergence in the iteration domain for the linearized dynamics and validate the performance by simulating the nonlinear dynamics.

eess.SY↗

arXiv2020

Impact of Network Topology on the Stability of DC Microgrids

We probe the stability of Watts-Strogatz DC microgrids, in which droop-controlled producers and constant power load consumers are homogeneously distributed and obey Kirchhoff's circuit laws. The concept of survivability is employed to evaluate the system's response to Dirac delta voltage perturbations at single nodes. A fixed point analysis of the power grid model yields that there is only one relevant attractor. Using a set of simulations with random networks we investigate correlations between survivability and three topological network measures: the share of producers in the network and the degree and the average neighbour degree of the perturbed node. Depending on the imposed voltage and current limits, the stability is optimized for low node degrees or a specific share of producers. Based on our findings, we provide an insight into the local dynamics of the perturbed system and derive explicit guidelines for the design of resilient DC power grids.

physics.soc-ph↗

arXiv2020

A multilayer, multi-timescale model approach for economic and frequency control in power grids

Power systems are subject to fundamental changes due to the increasing infeed of decentralised renewable energy sources and storage. The decentralised nature of the new actors in the system requires new concepts for structuring the power grid, and achieving a wide range of control tasks ranging from seconds to days. Here we introduce a multiplex dynamical network model covering all control timescales. Crucially, we combine a decentralised, self-organised low-level control and a smart grid layer of devices that can aggregate information from remote sources. The safety-critical task of frequency control is performed by the former, the economic objective of demand matching dispatch by the latter. Having both aspects present in the same model allows us to study the interaction between the layers. Remarkably, we find that adding communication in the form of aggregation does not improve the performance in the cases considered. Instead, the self-organised state of the system already contains the information required to learn the demand structure in the entire grid. The model introduced here is highly flexible, and can accommodate a wide range of scenarios relevant to future power grids. We expect that it is especially useful in the context of low-energy microgrids with distributed generation.

nlin.AO↗