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

Lia Strenge, Paul Schultz, Jürgen Kurths, Jörg Raisch, Frank Hellmann

Abstract

Power systems are subject to fundamental changes due to the increasing infeed of decentralized renewable energy sources and storage. The decentralized 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 decentralized, self-organized 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 and 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-organized 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.

Introduction

Traditional power grids rely on massive, centralized power plants to provide rotational inertia and centralized control. As energy systems transition toward decentralized renewable sources like wind and solar, this inherent stability is lost. Future "prosumer" microgrids must simultaneously address two challenges across vastly different timescales: maintaining frequency stability (seconds) to prevent equipment damage and achieving economic dispatch (hours to days) to balance supply and demand cost-effectively. While these problems are typically studied in isolation, this paper introduces a unified model to explore the interactions between fast-acting physical dynamics and slow-acting economic optimization in smart grids.

Methodology: The Multiplex Network Approach

The researchers developed a multiplex dynamical network model consisting of two distinct layers. The Physical Layer represents the actual power flow and frequency dynamics, modeled using Kuramoto oscillators with inertia. It employs a decentralized "leaky integrator" control for immediate frequency stabilization. Above this sits the Control/Smart Layer, which utilizes Iterative Learning Control (ILC). ILC is an adaptive algorithm that analyzes the control effort expended in previous cycles to schedule more efficient power dispatch for the current day.

The multiplex network model consists of a physical grid layer (bottom) and a smart control layer (top) that can aggregate information.

To test the model, the authors used synthetic demand curves combining a periodic daily baseline with stochastic Gaussian fluctuations. They simulated these dynamics across an ensemble of 100 random 3-regular graphs, each containing 24 nodes, using the Julia programming language. Five different control scenarios were evaluated, ranging from a baseline with no smart dispatch to fully connected communication networks.

Key Insights and Results

A primary finding of the study is that explicit data communication between smart controllers does not significantly enhance economic performance. The physical grid itself acts as a communication medium; demand changes cause frequency shifts that local controllers react to, and the ILC algorithm can "read" these reactions to infer the demand structure of the entire grid.

Furthermore, the study identified an optimal density for smart controllers. It is not necessary for every node to be "smart" to achieve system-wide efficiency.

Overall system cost as a function of the number of ILC-equipped nodes, showing an optimal performance point at approximately one-third of the nodes.

As shown in the analysis, the system reaches a cost-optimal state when approximately one-third of the nodes (8 out of 24) are equipped with ILC controllers. Beyond this point, adding more smart nodes provides diminishing returns or even slightly increases costs due to potential overcompensation and conflicting actions between decentralized controllers.

Performance Metrics

  • Cost Efficiency: The implementation of ILC reduced total system costs by a factor of four compared to grids relying solely on frequency control without economic dispatch.
  • Stability Gains: Although ILC is primarily an economic tool, it improved frequency stability by suppressing the demand fluctuations that typically drive the system out of equilibrium.
  • Communication Redundancy: Scenarios with structured communication performed almost identically to those with purely local smart controllers, confirming that the self-organized state of the grid is a sufficient information carrier for optimization.

Discussion and Future Work

While the model provides significant insights, it uses simplifications such as the Kuramoto model, which neglects voltage dynamics. Additionally, the ILC approach assumes a degree of periodicity in demand. Future research could extend this framework to include local constraints like battery storage limits and ramping times, or investigate using the ILC layer to prioritize green energy sources over fossil fuels. This flexible model is particularly promising for the design and management of low-energy microgrids with distributed generation.

Summary updated

Abstract

Power systems are subject to fundamental changes due to the increasing infeed of decentralized renewable energy sources and storage. The decentralized 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 decentralized, self-organized 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 and 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-organized 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.
BibTeX
@inproceedings{strenge2020multiplex,
  author = {Lia Strenge and Paul Schultz and Jürgen Kurths and Jörg Raisch and Frank Hellmann},
  title = {{A multiplex, multi-timescale model approach for economic and frequency control in power grids}},
  booktitle = {Chaos},
  year = {2020},
  doi = {10.1063/1.5132335},
}
Figure