Method and computer program product for determining grid bottlenecks for a power grid

Abstract

The invention relates to a method (100) for determining grid bottlenecks for a power grid, wherein the power grid comprises a plurality of components, in particular lines, switches and / or systems. The method is characterised by the following steps: - (102) providing a plurality of grid topologies for the power grid; - (104) carrying out an overload calculation for each of the grid topologies; - (106) determining an overload probability for each of the components by means of the grid topologies and the overload calculation; wherein - (108) the grid bottleneck is determined by determining at least one of the components which, when the component is assumed to be overload-free, leads to the greatest number of overload-free grid topologies. The invention further relates to a computer program product.

Description

[0001] 202505021

[0002] 1

[0003] Description

[0004] Method and computer program product for determining network bottlenecks for a power grid

[0005] The present invention relates to the determination of network bottlenecks for power grids, and in particular to a method for determining network bottlenecks by analyzing multiple network topologies and overload probabilities according to claim 1 and a computer program product according to claim 15.

[0006] In the field of electricity distribution, managing grid congestion plays a crucial role in ensuring an efficient and reliable power supply. Electricity grids, particularly distribution networks, face increasing technical challenges due to the growing integration of distributed energy resources (DERs), such as photovoltaic systems, electric vehicle charging stations, and heat pumps.

[0007] These developments lead to new complexities regarding maintaining network stability and preventing overloads.

[0008] Distribution system operators (DSOs) are typically responsible for proactive expansion planning in medium-voltage and low-voltage networks and must address the aforementioned technical challenges. To identify and determine network congestion, network areas can be identified that are likely to experience the greatest bottlenecks in future operating scenarios. This identification can then be used to prioritize physical expansion planning or to implement active congestion management measures.

[0009] Typically, network expansion planning in distribution systems is based on a single network topology determined by a fixed switching configuration during operation. However, this approach has drawbacks, as the standard switching configuration for a power grid can change over time, for example, due to maintenance work or the addition of new consumers. The large number of switches in medium-voltage or low-voltage networks leads to numerous possible configurations, resulting in several potential operating topologies for a power grid.

[0010] Furthermore, changes in switching configurations are typically poorly documented. This makes it difficult to determine the current network topology with certainty. This uncertainty complicates the process of evaluating expansion options.

[0011] 2

[0012] a basic topology. Therefore, there is a growing need for more robust planning methods that can consider multiple alternative network topologies.

[0013] By considering various network topologies during the planning phases, distribution system operators (DSOs) can identify which bottlenecks can be resolved through reconfiguration and which must be prioritized for network expansion or further network interventions. This ensures that expansion measures not only resolve bottlenecks for a single planning topology or network topology, but also for other switching configurations / network topologies that may be required in the future. Therefore, it is necessary to consider various possible operating topologies of electricity networks during network expansion and bottleneck management.

[0014] The present invention is based on the objective of providing an improved method for determining or identifying network bottlenecks for a power grid, which in particular takes into account several possible network topologies of the power grid.

[0015] The problem is solved by a method with the features of independent claim 1 and by a computer program product with the features of independent claim 15. Advantageous embodiments and further developments of the invention are specified in the dependent claims.

[0016] The inventive method for determining network bottlenecks (determining one or more network bottlenecks) for a power grid, wherein the power grid comprises several components, in particular lines, switches and / or systems, is characterized by at least the following steps:

[0017] - Providing multiple network topologies for the power grid;

[0018] - Performing an overload calculation for each of the network topologies;

[0019] - Determining an overload probability for each of the components and for each network topology using the network topologies and the overload calculation; wherein

[0020] - the network bottleneck determination is carried out by identifying at least one of the components which, if the component is assumed to be free of overload, leads to the largest number of overload-free network topologies.

[0021] A grid bottleneck can refer to a state of the power grid in which at least one of its components is overloaded, that is, at least one.

[0022] 3

[0023] Network-specific and / or component-specific limits have been violated. For example, this could be a voltage band violation and / or a power limit violation. A component is considered free of overload if it does not exhibit an overload.

[0024] A network topology can refer to the physical and electrical configuration of a power distribution network, including the arrangement of lines, switches, transformers, and / or other components. Several network topologies are considered here to account for various possible network configurations that may arise from switching operations or network reconfigurations.

[0025] An overload calculation is used to determine, for a given network topology, which components of the power grid are experiencing an overload. This overload calculation can, in particular, be a load flow calculation.

[0026] An overload probability is a statistical measure representing the likelihood of a component in a network topology experiencing an overload. The overload probability is determined based on the results of overload calculations performed for each network topology. An aggregated overload probability considers the overload probabilities across all network topologies. This can be calculated, for example, by averaging or weighting the individual overload probabilities across the network topologies.

[0027] A component can be an element of the power grid and / or the associated electrical infrastructure. A component can be designed as a line, in particular as a transmission and / or distribution line, as a switch for grid reconfiguration, and / or as a plant, in particular as a transformer, as a local network substation, as a substation, and / or as a distributed energy resource.

[0028] The invention has in particular one or more of the advantages mentioned below:

[0029] Consideration of multiple network topologies: In contrast to known methods that rely on a single network topology, the invention considers multiple network topologies. This approach takes into account the dynamic nature of modern power grids, where 202505021

[0030] 4

[0031] Switching configurations can change due to maintenance work, new consumer connections, and / or operational requirements. By analyzing various possible network configurations, the method provides a more robust and realistic identification of potential bottleneck problems.

[0032] Probabilistic bottleneck identification: By calculating the probability of overload, network planners can prioritize network interventions based on statistical evidence and do not have to rely on worst-case scenarios or single-point estimates.

[0033] Identifying critical components: This method determines which components, assuming they are free of congestion, result in the greatest number of congestion-free network topologies. This symbolically identifies the components most influential in reducing bottlenecks. As a result, network operators can optimize expansion strategies and achieve more effective bottleneck relief through targeted network interventions.

[0034] Improved planning for future grid configurations: With the increasing integration of distributed energy resources, considering multiple grid topologies is becoming ever more important. The invention enables grid planning to assess the impact of potential future grid configurations and to ensure that expansion measures and congestion management strategies remain effective and sufficient across different scenarios.

[0035] Improved reliability and resilience: By considering multiple network topologies, bottleneck problems that may occur under different network configurations can be identified. This increases the reliability and resilience of the power grid by taking potential bottlenecks into account in various operating conditions.

[0036] Efficient resource allocation: By identifying the component that has the greatest impact on congestion-free network topologies, as defined by the present invention, more efficient network control and improved congestion management are enabled. Network operators can prioritize the identified component(s) during network expansion, thereby reducing overall costs and simultaneously increasing the effectiveness of congestion management measures.

[0037] 5

[0038] Adaptability to complex network structures: Modern power grids, particularly at the distribution level, are becoming increasingly complex due to the integration of DERs, smart grid technologies, and flexible loads. The method according to the invention is suitable for managing this complexity by analyzing multiple network topologies, thereby providing a holistic overview of bottleneck risks.

[0039] Support for active grid management: As power grids transition to more active management approaches, the ability to quickly assess bottlenecks across multiple configurations is becoming increasingly important. This invention provides a basis for developing dynamic bottleneck management strategies that can adapt to changing grid conditions.

[0040] Improved decision-making for network expansion: By providing insights into which components have the greatest impact on reducing bottlenecks across multiple network topologies, the invention supports more informed decision-making in network expansion planning. This can lead to more cost-effective and future-proof infrastructure investments.

[0041] Compatibility with existing planning processes: The invention can be integrated into existing network planning workflows and improves current practices without requiring a complete overhaul of established procedures. This compatibility facilitates introduction and implementation by distribution network operators.

[0042] The computer program product according to the invention is characterized in that it comprises instructions which, when executed by a computing unit, in particular a computer, cause it to execute a method and / or steps of the method according to one of claims 1 to 14.

[0043] The inventive method for determining the network topology offers similar, equivalent and equivalent advantages.

[0044] According to a preferred embodiment of the invention, the determination of at least one component is carried out by means of an optimization method.

[0045] This enables efficient identification of the critical components that, in accordance with the invention, have the greatest impact on reducing network bottlenecks.

[0046] 6

[0047] This approach is particularly advantageous for complex power grids with a multitude of components and / or network topologies. By using optimization methods, a large number of potential solutions can be evaluated, and the optimal group of components can be identified for prioritization in bottleneck reduction.

[0048] According to a preferred embodiment of the invention, in the optimization method an objective function is maximized, wherein the objective function models the number of overload-free network topologies.

[0049] Similarly, an objective function can be used that is minimized. This objective function represents a measure of the number of congestion-free network topologies. This number depends on the respective components, so maximizing the objective function or its value identifies the components that lead to the maximum number of congestion-free network topologies. This allows at least one component, or a potentially defined number of components, to be identified that, if assumed to be congestion-free (for example, by expanding the respective component), will result in the greatest number of congestion-free network topologies.

[0050] According to a preferred embodiment of the invention, the network bottlenecks are determined by the fact that

[0051] - in a first step, the number of overload-free network topologies is determined for each of the components, whereby the respective component is assumed to be overload-free;

[0052] - in a second step, one of the components is identified that leads to the largest number of overload-free network topologies among the components or has the highest aggregated overload probability; and

[0053] - the first and second sub-steps are repeated, whereby the component determined in the previous second sub-step is not taken into account in each repetition.

[0054] This iterative approach provides a systematic method for identifying and prioritizing components that have the greatest impact on reducing network congestion. By sequentially determining the component that leads to the greatest number of congestion-free network topologies or exhibits the highest aggregated probability of congestion, the method ensures that each step contributes to the most optimal improvement in network performance.

[0055] 7

[0056] This leads to an efficient procedure that results in a sufficiently accurate determination of the components.

[0057] According to a preferred embodiment of the invention, the partial steps are repeated until a specified number of components has been determined or no more overloads are present.

[0058] This termination condition offers flexibility in the application of the method and allows for adaptation to specific planning needs or resource constraints. By defining a determined number of components, network operators can focus on a fixed number of priority network interventions. Alternatively, continuing until all overloads disappear ensures a comprehensive solution for network congestion, but potentially at a higher cost or with greater complexity. This approach allows for a balance between congestion reduction and practical implementation limitations.

[0059] According to a preferred embodiment of the invention, the identified components are provided by means of a sorted list, wherein the arrangement of the components within the list is associated with a prioritization with regard to network interventions.

[0060] Outputting the results as a sorted list provides network planners and operators with a clear and actionable report. The prioritized list enables efficient resource allocation by focusing on the components that have the greatest impact on reducing bottlenecks. This approach facilitates strategic decisions regarding network expansion and / or component upgrades.

[0061] According to a preferred embodiment of the invention, a network topology is overload-free if the overload probability of each of the components has a value less than or equal to a threshold value.

[0062] Defining an overload-free topology based on a probability threshold offers a flexible and improved approach to bottleneck assessment. This method acknowledges that a certain level of overload risk can be acceptable in practical network operation and allows for a balance between reliability and cost-efficiency. The threshold can be set based on specific network requirements or risk tolerance levels. Furthermore, the overload probability can take only the discrete values ​​of zero or one.

[0063] 8

[0064] The component is unloaded for a value of zero and overloaded for a value of one according to the overload calculation.

[0065] According to a preferred embodiment of the invention, when determining network bottlenecks, only components are taken into account that have an overload probability greater than the threshold value, in particular greater than zero.

[0066] This approach focuses the analysis on components most likely to contribute to network bottlenecks. By excluding components with very low or zero overload probabilities, complexity is reduced and more efficient use of computational resources is enabled. This is particularly advantageous for large-scale network systems with numerous components.

[0067] According to a preferred embodiment of the invention, the overload probability of a component is determined based on the number of network topologies for which the respective component exhibits an overload according to the respective overload calculation.

[0068] This method for calculating the probability of congestion provides a simple and efficient measure of a component's susceptibility to bottlenecks across multiple network configurations. By considering the proportion of network topologies in which a component exhibits congestion, the approach captures the component's vulnerability to bottlenecks under various operating conditions. This probabilistic assessment offers a more accurate view of bottleneck risk compared to deterministic methods.

[0069] According to a preferred embodiment of the invention, the overload calculation is performed by means of a load flow calculation and / or by means of an aggregation.

[0070] The use of load flow calculations offers an advantageous method for determining the overload of a component, taking into account the respective network state. This approach allows for the consideration of the complex interactions between different components and the impact of load flow distributions on bottlenecks. The use of aggregation methods provides flexibility when handling large systems or situations where detailed simulations are too computationally intensive.

[0071] 9

[0072] According to a preferred embodiment of the invention, several operating topologies of the power grid are provided as network topologies, wherein the operating topologies are formed by means of different positions of switches of the power grid.

[0073] Considering multiple operating topologies, formed by different switch positions, enables a comprehensive analysis of potential grid configurations. This approach takes into account the flexibility inherent in modern, complex power grids. By analyzing various possible switch configurations, the method provides insights into how changes in grid topology can affect bottleneck patterns.

[0074] According to a preferred embodiment of the invention, a Markov Chain Monte Carlo algorithm is used to determine possible operating topologies.

[0075] The use of a Markov Chain Monte Carlo (MCMC) algorithm to identify possible operating topologies provides an efficient method for sampling from the space of potential network configurations. This approach enables the identification of a technically representative set of network topologies without requiring an exhaustive enumeration, which would be computationally unfeasible for large power grids. MCMC methods are particularly well-suited for this task because they can handle complex, high-dimensional spaces and can be configured to focus on more probable, and therefore more technically relevant, configurations.

[0076] According to a preferred embodiment of the invention, the components are designed as lines, switches, substations and / or local network stations.

[0077] This enables a holistic analysis of bottlenecks across different levels and elements of the power distribution system. By including lines, switches, substations, and / or local distribution substations, the method can identify bottleneck hotspots and critical components throughout the entire network hierarchy. This approach ensures that bottleneck problems are addressed at multiple levels, from individual lines to large substations.

[0078] According to a preferred embodiment of the invention, the power grid is configured as a medium-voltage or low-voltage power grid.

[0079] 10

[0080] Focusing on medium- and low-voltage power grids addresses a critical area of ​​the electricity distribution system where bottlenecks are increasingly occurring. These voltage levels are particularly affected by the integration of distributed energy resources and changing consumption patterns. By adapting the methodology to these grid segments, it provides valuable insights for distribution system operators facing new technical challenges in grid management.

[0081] Further advantages, features, and details of the invention will become apparent from the exemplary embodiments described below and from the drawings. These schematically illustrate:

[0082] Figure 1 shows a flowchart for a procedure for identifying network bottlenecks in an electricity grid;

[0083] Figure 2 shows a matrix for determining network bottlenecks in a power grid; and Figure 3 shows a power grid with several components, in particular network nodes and lines.

[0084] Similar, equivalent, or equivalent elements can be provided with the same reference symbols in the figure.

[0085] Figure 1 shows a flowchart for a method 100 for determining network bottlenecks in a power grid. Method 100 comprises several steps 102, 104, ..., 108, which analyze multiple network topologies to identify bottleneck hotspots and prioritize components for potential network interventions. For this purpose, the respective components with the greatest impact on overload-free network topologies are determined.

[0086] In step 102, several network topologies are provided for the power grid. This takes into account various possible network configurations that arise, for example, for different switch positions. These network topologies can thus represent different switching states, planned expansions, and / or potential future scenarios. The generation of the network topologies could be achieved through various means, such as historical data analysis, expert knowledge, and / or algorithmic generation, in particular using a Markov chain Monte Carlo algorithm.

[0087] Step 104 involves performing an overload calculation for each of the network topologies provided in Step 102. This calculation typically includes performing 202505021

[0088] 11

[0089] Load flow simulations are performed for each network topology under various load and generation scenarios. These simulations determine whether components of the power grid, such as lines, transformers, substations, and / or local distribution substations, exceed their rated capacities and / or cause a voltage band violation under the given conditions.

[0090] In step 106, an overload probability is determined for each component and for each network topology in the power grid. The overload probability is calculated based on the results of the overload calculations for all analyzed network topologies. The overload probability provides a quantitative measure of how likely a component is to experience bottlenecks across different network configurations. This probabilistic approach offers an improved understanding of bottleneck risks compared to known deterministic methods.

[0091] In step 108, network bottlenecks are identified by pinpointing at least one component. This involves determining which component, assuming it is free of overload, leads to the greatest number of overload-free network topologies. This step is advantageous for prioritizing network interventions in the power grid. It advantageously identifies the component with the greatest technical impact on preventing network bottlenecks. This component should then be prioritized with regard to control / regulation measures and / or network expansion.

[0092] Figure 2 shows a matrix for determining network bottlenecks in a power grid and illustrates the procedure described in Figure 1.

[0093] Columns 201 of the matrix shown correspond to various components of the power grid. These components can be, in particular, lines, switches, transformers, substations, and / or local distribution substations. Rows 202 correspond to several possible network topologies of the power grid. These network topologies represent different possible configurations of the power grid and take into account, in particular, different switching states, planned expansions, and / or potential future scenarios.

[0094] The matrix shown depicts unloaded components 206 (without hatching) and overloaded components (with hatching) 204. For clarity, only exemplary components are marked with reference numbers 204 and 206, respectively. (202505021)

[0095] 12

[0096] A component is considered free of overload if it does not exhibit an overload for the respective network topology (rows 202) according to the respective overload calculation. Similarly, a component is considered overloaded if it exhibits an overload for the respective network topology (rows 202) according to the respective overload calculation. The matrix shown thus indicates which components exhibit an overload (hatching) and therefore a network bottleneck for each network topology. A network topology therefore has a bottleneck if one or more of its components are overloaded for that network topology. An overload-free network topology comprises only overload-free components; that is, all components in the respective network topology are overload-free.

[0097] The matrix shown allows for the visualization of bottleneck patterns across multiple network topologies. This makes it possible to identify the component that has the greatest impact on network bottlenecks, or, assuming no overload, the component that leads to the greatest number of overload-free network topologies. The aggregated overload probability of a component can be determined by the proportion of network topologies in which the component exhibits an overload.

[0098] The matrix illustrates that, for example, expanding and / or upgrading component j so that it is overload-free in all network topologies does not result in an overload-free network topology, as other components remain overloaded. Expanding and / or upgrading component i so that it is overload-free, or can be assumed to be overload-free, leads to at least one overload-free network topology; namely, network topology 1 would then be overload-free. Therefore, according to the example matrix, the expansion and / or upgrade of component i should be prioritized.

[0099] Figure 3 shows a network diagram 300, which represents a medium voltage network and / or low voltage network with several interconnected components, network nodes and branches.

[0100] The figure provides a more detailed view of how the bottleneck analysis method can be applied to a specific network topology.

[0101] The power grid comprises several root buses 300, 302, and 304. In this example, root bus 300 is unloaded, and root bus 302 is overloaded. Typically, root buses 300, 302, and 304 are formed by local network substations, substations, and / or transformers of the power grid. At least one overloaded line of the 202505021

[0102] 13

[0103] The power grid is marked with the reference symbol 306. As an example, another root bus 306 is marked in Figure 3, which has an overload probability of approximately 50 percent.

[0104] Overload probabilities are determined using overload calculations and network topologies. For this purpose, the proportion of network topologies for which the respective component exhibits an overload according to the overload calculation is determined. Thus, for example, root bus 306 exhibits an overload in 50 percent of the network topologies. In the other 50 percent of the network topologies, root bus 306 is free of overload. The overload can preferably be determined using a load flow analysis. Should the load flow analysis not converge numerically, the power grid can be divided into subnetworks for which the load flow analyses can then be performed convergently.

[0105] Although the invention has been further illustrated and described in detail by the preferred embodiments, the invention is not limited by the disclosed examples, nor can other variations be derived from them by a person skilled in the art without departing from the scope of protection of the invention. 202505021

[0106] 14

[0107] Reference symbol list

[0108] 100 procedures

[0109] Step 102

[0110] Step 104

[0111] Step 106

[0112] 108 steps

[0113] 201 Columns (Component) 202 Rows (Network Topologies) 204 Overloaded Component 206 Overload-Free Component 300 Power Grid

[0114] 300 overload-free root bus 302 overloaded root bus 304 root bus

[0115] 304 Root Bus

[0116] 306 overloaded line

Claims

202505021 15 Patent claims 1. Method (100) for determining network bottlenecks for an electricity network, wherein the electricity network comprises several components, in particular lines, switches and / or installations, characterized by the following steps: - (102) Providing multiple network topologies for the power grid; - (104) Perform an overload calculation for each of the network topologies; - (106) Determining an overload probability for each of the components and for each network topology using the network topologies and the overload calculation; wherein - (108) the identification of network bottlenecks is carried out by identifying at least one of the components which, if the component is assumed to be free of congestion, leads to the largest number of free-of-congestion network topologies.

2. Method (100) according to claim 1, characterized in that the determination of the at least one component is carried out by means of an optimization method.

3. Method (100) according to claim 2, characterized in that in the optimization method an objective function is maximized, wherein the objective function models the number of overload-free network topologies.

4. Method (100) according to claim 1, characterized in that the network bottlenecks are determined by the fact that - in a first step, the number of overload-free network topologies is determined for each of the components, whereby the respective component is assumed to be overload-free; - in a second step, one of the components is identified that leads to the largest number of overload-free network topologies among the components or has the highest aggregated overload probability; and - the first and second sub-steps are repeated, whereby the component determined in the previous second sub-step is not taken into account in each repetition.

5. Method (100) according to claim 4, characterized in that the partial steps are repeated until a specified number of components has been determined or no more overloads are present. 16 6. Method (100) according to one of the preceding claims, characterized in that the identified components are provided by means of a sorted list, wherein the arrangement of the components within the list is associated with a prioritization with respect to network interventions.

7. Method (100) according to one of the preceding claims, characterized in that a network topology is overload-free if the overload probability of each of the components has a value less than or equal to a threshold value.

8. Method (100) according to claim 7, characterized in that, when determining network bottlenecks, only components are taken into account which have an overload probability greater than the threshold value, in particular greater than zero.

9. Method (100) according to one of the preceding claims, characterized in that the aggregated overload probability according to claim 4 of a component is determined based on the number of network topologies for which the respective component has an overload according to the respective overload calculation.

10. Method (100) according to one of the preceding claims, characterized in that the overload calculation is carried out by means of a load flow calculation and / or by means of an aggregation.

11. Method (100) according to one of the preceding claims, characterized in that several operating topologies of the power grid are provided as network topologies, wherein the operating topologies are formed by means of different positions of switches of the power grid.

12. Method (100) according to claim 11, characterized in that a Markov Chain Monte Carlo algorithm is used to determine possible operating topologies.

13. Method (100) according to one of the preceding claims, characterized in that the components are configured as lines, switches, substations and / or local network stations. 17 14. Method (100) according to one of the preceding claims, characterized in that the power grid is designed as a medium-voltage or low-voltage power grid.

15. Computer program product comprising instructions which, when executed by a computing unit, in particular a computer, cause it to execute a method and / or steps of the method (100) according to any one of claims 1 to 14.

Images