As the number of IBR sources in the GB system continues to rise, it becomes essential to conduct numerous Electromagnetic Transient (EMT) simulations across various scenarios and contingency cases to assess system stability. Consequently, there is an escalating demand for the development of the ability to perform multiple EMT analyses to facilitate broader network studies while keeping simulation times manageable and practical.
This project aims to develop innovative approaches for expediting simulation times required to execute the comprehensive GB EMT model. It will also offer technical insights to ascertain the imperative need for EMT simulations during critical system conditions.
Benefits
The project will enhance the GB network's EMT model by improving the models' computational efficiency, which will help investigate more scenarios with stability risks while transitioning into zero carbon operation. Additionally, it will offer technical recommendations to identify critical system conditions requiring EMT simulations.
The learnings from this project will also be beneficial to Transmission Owners (TOs) concerning the run time of their respective EMT networks. The TOs use the same EMT software package, and the developed tool should be able to integrate with their models seamlessly. Furthermore, the second phase of the innovation project will produce technical guideline outlining the scenarios for which EMT simulations are necessary under system-critical conditions for better-informed decisions.
Learnings
Outcomes
The project delivered:
• A fully enhanced and optimised PSCAD model of the E&W transmission network, including inclusion of at least 40 generic representations of HVDC embedded and interconnector links, offshore wind farms (AC and DC connected) and STATCOMs, that runs in a practical runtime (circa. 20 minutes from several hours) for 30 seconds simulation on multi‑core hardware. All additional inverter based models were required to demonstrate a practical runtime before being integrated into the enhanced model.
• A Technical Guide detailing when EMT studies are necessary, including qualitative phenomena and quantitative thresholds, and providing a repeatable workflow for boundary selection and network reduction.
• Workflows and scripts/tool based on impedance scans and voltage dip in PowerFactory that allow the system operator to quickly identify the network boundary to retain in EMT when deciding the region to reduce for system level studies.
• Stakeholder knowledge transfer, including workshops, reports and example studies, enabling NESO engineers to apply the methods in future connection assessments and planning studies.
These outcomes equip NESO with the capability to conduct wide‑area EMT studies efficiently and consistently and to make informed decisions on when EMT is required.
Lessons Learnt
1. Early runtime profiling was useful.
Wide-area EMT models were found to exhibit severe numerical stiffness and performance bottlenecks. Profiling and partitioning were undertaken early to avoid wasted effort on unworkable configurations.
2. Generic models still needed to capture key controls and vendor-level behaviour.
Even when proprietary detail could not be shared, generic EMT models needed to include representative control loops (current limiting, voltage control, PLL) to produce meaningful results. Over-simplified models were found to risk misleading conclusions.
3. Boundary selection required both RMS and EMT insight.
RMS studies were useful for screening candidate boundaries, but frequency-dependent impedance analysis was required to ensure the reduced external network replicated the dynamic behaviour seen by converter controls. A combined approach improved confidence in results.
4. Stakeholder engagement improved guidance usability.
Workshops with users helped to refine screening thresholds, clarify terminology and ensure that the Technical Guide was accessible to non-specialists. Early and regular engagement was recommended.
5. Flexible workflows supported scaling.
Scripts and automation for case preparation, initialisation and runtime management were invaluable when dealing with many partitions and scenarios. These were designed to accommodate model updates and additional future connections.
