Zukunftslabor
ENERGIE

The integration of renewable energy sources and the increasing demands from the electrification of transportation and heating introduce significant variability and uncertainty to distribution grids. This variability can lead to operational challenges, including voltage range violations and instability in power supply. To address these challenges, it is crucial to develop robust methodologies that quantify uncertainties and evaluate their impacts on grid parameters, thereby informing decision-making processes to enhance the reliability and efficiency of Active Distribution Networks (ADNs).

Leveraging the enhanced measurement and data acquisition capabilities from the smart meter rollout in Germany and the network-oriented control regulations outlined in §14a of the Energy Industry Act (EnWG), this research presents a configurable framework for quantifying uncertainties and measuring their impacts on grid operational parameters. This framework supports risk assessments to identify potential violations of permissible limits for these grid parameters, aiding critical grid operation decisions.

The framework employs open-source data and libraries to ensure transparency, reproducibility, and accessibility, facilitating comprehensive grid simulations. Utilizing a real residential district in Lower Saxony as a model, the grid includes 49 buildings equipped with PV systems, EV charging stations, and battery storage systems. Key components of the framework include libraries such as pandapower for low-voltage grid simulation, PVlib for PV generation modelling, emobpy for EV charging profiles, and weather data from the German Weather Service. The primary source of uncertainty in this study is the error in Global Horizontal Irradiance (GHI) affecting PV generation, though the framework is adaptable to other forecasted parameters.

The methodology involves calculating forecast errors, performing statistical analysis, and generating random samples to model PV generation at each bus. These samples are integrated into a quasi-dynamic grid simulation to evaluate the sensitivity of voltage range violations to the identified uncertainties. Monte Carlo simulations are used throughout this process to provide a probabilistic assessment of potential grid stability issues, supporting informed decisions on power injection and grid management based on voltage sensitivity analysis. By adjusting the uncertainty variables, this framework facilitates the investigation of various factors to determine their impact on grid stability. To isolate the effects of the forecast variable under investigation, other variables are set to their actual values. This method identifies critical issues and pinpoints the most vulnerable buses in the grid, enhancing the understanding of grid dynamics under varying conditions and aiding in the development of targeted risk mitigation strategies.

Future work aims to enhance the reliability and efficiency of ADNs by integrating Uncertainty Quantification with advanced control strategies. This includes developing dynamic mechanisms for proactively detecting and mitigating limit violations in key grid parameters through sensitivity analysis and optimized power injection. The goal is to improve grid stability and operation while providing valuable insights into uncertainty propagation within grid systems.

The dynamic nature of power systems complicates future low voltage (LV) grid operations. Variable power demand, fluctuating renewable energy, and expanding electric heating and mobility pose significant challenges for grid management. To avoid grid expansion, effective operating strategies are crucial to prevent issues from the lack of real-time coordination between distributed energy resources and grid capacity. Flexible and proactive management enables operators to monitor and control critical grid parameters, enhancing performance and reliability.

Simulation models are essential for detailed energy system analysis, considering diverse load scenarios and supporting management strategy development. Co-simulations, integrating models from different domains, are particularly effective for thorough system analysis. These simulations should incorporate flexible and compatible algorithms, facilitating extensive testing and evaluation across a wide range of scenarios.

Existing research on centralized grid management has been extensively explored in high voltage systems but is sparse for LV grids. Studies often focus on voltage control, overlooking or simplifying line current issues, e.g. by assuming DC power flow to narrow the focus on the applied regulation strategies or to avoid complexity associated with an accurate analysis. This paper presents an enhanced algorithm in an open-source tool to identify and correct voltage and line loading violations in LV grid simulations. The tool offers configurable options for tailoring power management strategies and supports selective injections of reactive and active power using various sources like battery storage and photovoltaic (PV) systems. It employs a sensitivity matrix to accurately capture the dependencies of line currents on power injections at grid buses, ensuring precise modelling and improving convergence rates. The performance is compared with DC power flow models using simplified PTDF matrices to evaluate the necessity of comprehensive sensitivity analysis under varying grid dynamics. Additionally, different sequences for managing violations of operation limits are assessed to ensure gradual changes in grid parameters and avoid unnecessary power usage by carefully managing power injections. After validation, the tool is integrated into a co-simulation framework of the entire energy system of a district grid in Germany, enhancing testing and strategy assessment for grid performance. This builds on prior research analysing the impact of electric vehicle (EV) penetration on grid capacity, focusing on user-side measures like storage and smart charging to manage voltage issues. The co-simulation integrates models and data, including actual grid simulation, DWD weather data, measured load profiles, and models for PV generation and EV charging, ensuring comprehensive and realistic energy system analysis.

The growing complexity of smart grids, driven by distributed energy systems and electric vehicles, poses challenges in managing power flow and operational boundaries. This study explores using storage systems as a flexibility tool in smart grids. A case study in northern Germany models a smart grid with energy generation, use, and storage components. A co-simulation framework integrates these models with Home Energy Management Systems (HEMS), balancing energy across buildings. A central control agent monitors grid voltage, while flexibility agents optimize operational plans by communicating with the HEMS. The study evaluates scenarios with and without central and flexibility optimisation. The inclusion of central control and flexibility optimization reduced the amount of voltage range violation incidents by 81%, demonstrating the system’s effectiveness. Moreover, the mean value of state of charge of the batteries is reduced from 42% to 30%, indicating a reduced storage demand. Future work will expand the control system to address line loading limitations and explore electric vehicle charging strategies. The integration of other storage systems, such as thermal storage, could further optimise energy management in the district, leading to a comprehensive multi-domain flexibility control approach.