LINZ NETZ and GridData evaluate the DigitalTwin
Measurement-based grid connection evaluation with a digital twin. Over 17,000 installed PV systems and 288,000 connection points at LINZ NETZ - these are impressive figures that clearly show how important it is to efficiently handle the ever-increasing demands on grid flexibility. This is further reinforced by the connection of new consumers such as heat pumps or charging stations. Various options are available to meet these challenges, such as conventional grid expansion, reactive power management or controllable local grid transformers. A meaningful grid calculation is crucial for the correct selection of measures.
Enhancing grid flexibility with probabilistic analysis
The "conventional" grid calculation widely used today is based on worst-case assumptions, which often make grid expansion necessary (equipment, cable costs, personnel deployment). A so-called "probabilistic" approach works with expected values, e.g. the voltage at the station transformer or the PV feed-in power. The voltage must not exceed limit values with a high probability. A measurement-based approach can be helpful to prove the high probability. For this purpose, quarter-hourly measurements of active and reactive power or voltage from smart meters, load profile meters, meters in transformer stations and PV inverters are used in time series-based load flow calculations. This is expected to provide a realistic representation of the grid conditions that can be used for the probability assessment.
Probabilistic Modeling
DigitalTwin Technology
A pilot project by LINZ NETZ and GridData
LINZ NETZ and GridData evaluated this approach in a joint pilot project in 2022/2023. For this purpose, a limited grid area was selected and the topological and available measurement data was mapped in GridData DigitalTwin. The objectives of the project were:
- Develop an understanding of the possible applications of a digital grid twin
- Assessing the potential of a measurement-based grid connection evaluation
- Testing the functions of the GridData DigitalTwin software
- Derive development priorities from the use in DSO operation
Coordinated approach - Trustful collaboration
Constraints in implementing the approach with a digital twin:
- Network security must be ensured.
- Established processes cannot be radically replaced.
- Which processes are affected by the new approach requires clarification.
With the following implications for collaboration:
- Incremental approach.
- Gain experience with new approaches.
- Gather feedback from departments and process owners regarding the implications of a changed approach.
The implementation steps were as follows:
- Install and familiarize with DigitalTwin.
- Select a suitable, delimited low-voltage grid section.
- Collect topology data and measurement data or metadata of the selected grid section and import, integrate and combine them accordingly in the DigitalTwin.
- Set up cloud-based access to the software and the option to visualize the time series of measurement data in the grid topology.
- Extend the pure monitoring of measured data to include the option of time-series-based load flow calculation. This requires the estimation of unmeasured grid connections.
- Evaluation of the calculation results of the time series-based load flow calculations in the connection evaluation, which inevitably occur due to the estimation of unmeasured grid connections. The accuracy is determined by comparing calculated and measured voltage time series.
- Consideration of topological changes over time - e.g. the integration of an additional transformer station
-
Development of possible future functional enhancements, especially in the DigitalTwin usage model.
At LINZ NETZ, for example, around 30 employees are involved in evaluating connection applications for PV systems.
Project collaboration:
- There was a great willingness to cooperate and learn together on both sides.
- The project involved few on-site meetings and regular (bi-weekly) online meetings.
- GridData employees were easily reachable, allowing pending questions to be quickly resolved.
- GridData's consistently positive attitude was evident in their flexible response to requirements from the analyses.
Analysis shows: Measurement-based methods are safe and accurate.
The accuracy of digital grid mappings is a recurring question. Based on the pilot implementation with the GridData DigitalTwin, it can be clearly stated that a good accuracy of the calculated results can be achieved. The difference between measured relative voltage values (based on nominal voltage) and calculated relative voltage values is around 1%. This deviation is due to the relatively small number of measured grid connections in the selected grid map (approx. 10%) and increases with each additional measurement taken into account. The foreseeable trend towards comprehensive quarter-hourly measurements with smart meters and the possibility of using this data for the purpose of expansion planning by grid operators (buzzword ElWG) will result in a further increase in the achievable accuracy.
How was the accuracy of the network model determined?
A basic amount of measurement data was available during the calculations (partly voltage measurement data) that could be used for plausibility checks. A comparison of the model data with the measurement data showed an inaccuracy of around 1 % (in relation to the nominal voltage). Due to the possible inaccuracy, a reserve of 1% must be included in the voltage band of +-10% to be observed by DSOs. Similarly, a further reserve of 2% must be included for voltage unbalance, which must be taken into account due to a lack of corresponding measurement data or due to the (symmetrical) calculation algorithm used. Although no large-scale analysis was carried out in the pilot project - the inaccuracy can vary somewhat due to special cases - the applicability and accuracy of digital grid presentations (here with GridData DigitalTwin) could be demonstrated.
- The intensive analysis of the achievable accuracy was important:
- As the safety of the power supply must be guaranteed under all circumstances, confidence in the quality of the measurement-based grid analysis is essential for its continued use.
- It should be noted that the high quality of the calculated data could be proven despite reduced measurement coverage.
Data Validation
High value for DSO confirmed when transitioning from worst-case to time-series measurements
An approach using measurement data is based on the consideration of load and feed-in time series. In conventional (worst-case) considerations, on the other hand, points in time are considered which represent the worst case. In the connection assessment of PV systems, for example, the maximum possible output of all feed-in systems is considered simultaneously in such a conventional approach, while the load curve is not taken into account. The power-reducing effect of "simultaneity" of load and generation is thus neglected. A measurement-based approach, on the other hand, takes this effect into account and allows a more precise insight into the actual grid conditions.
However, the maximum occurring voltage, which is used for the connection assessment, is provided with a probability in the measurement approach - a change in customer behavior can lead to a violation of voltage limits in the future. In order to ensure the voltage quality in such cases, it may be necessary to curtail generation plants. The number of such possibly necessary curtailments or shutdowns should be kept as low as possible and can be better estimated using a measurement-based approach (as requested by E-Control). However, further investigations are recommended.
In principle, the integration of measured values in grid operation, grid planning or connection assessments can be recommended based on the results of the study. For example, the need for flexible grid access in accordance with the Energy Industry Act or the flexible feed-in proposed by PV Austria can be investigated. The flexibility management required by the Energy Industry Act in Germany also makes it necessary to integrate real-time characteristics of the grid.
Accuracy of the DigitalTwin confirmed – project extensions available
From GridData's perspective, important insights for development priorities emerged during the project, based on feedback from LINZ NETZ GmbH. For instance, operational requirements such as managing multiple network models for comparative analysis, including switching history, were implemented during the project period. The provision of training videos addressed the need for offline training or self-learning opportunities. The visual representation of the network was enhanced by data channels to integrate with the existing process structure of the DSO.
Collaborative Development
The start has been made - digitization in network operations is progressing
The goals mentioned at the outset were clearly achieved thanks to the close collaboration within the team between LINZ NETZ GmbH and GridData:
- Develop an understanding of the applications of a digital grid twin
- Assessment of the potential of measurement-based network connection evaluation
- Testing the functions of the GridData DigitalTwin software
- Deriving development priorities from the application in DSO operation.
Many additional ideas for the use of digital grid twins have emerged – both from the perspective of network operations and software development.
For a widespread adoption of digital twins, an appropriate design of the basic systems (database, network connection, CPU performance) is necessary. Similarly, optimizing the measurement data used, especially in the low-voltage network, can support daily operations.
In this context, real-time capability is often mentioned. Real-time is always relative to the availability of measurement data in the power grid. Therefore, the focus should be on the informativeness of the available data for accurate network security prognosis. Further investigations are probably necessary for optimization.
Assistance in capturing the current network configuration with digital tools on-site can be helpful for connection assessments – this automatically synchronizes the network model in the DigitalTwin with past and current switching states.
For DSOs, there is a need to architecturally underpin the extended integration of the capabilities of a digital grid twin into the existing IT landscape, thereby unlocking further potential of digitization (e.g., with regard to asset management).
Participants:
- LINZ NETZ GmbH: G. Dobetsberger, H. Köpplmayr, J. Marchgraber
- Kontron Technologies: J. Zeindl, E. Detter
- GridData: H.-P. Schwefel, M. Reitenspieß
The RES Project Austria is supported by the German Federal Ministry for Economic Affairs and Climate Action as part of the Renewable Energy Solutions Programme of the German Energy Solutions Initiative.