Introduction
NESO is planning to introduce a new Grid Code modification, ‘GC0168 – Submission of Electromagnetic Transient (EMT) Models’. The aim of this modification is to require existing generators already connected to the network to provide an EMT model (in PSCAD) of their plant to NESO. While this may sound straightforward, it could potentially lead to significant issues and added costs that existing grid-connected generators should be aware of.
Background
The reason behind the GC0168 Grid Code modification is NESO’s goal to develop a comprehensive EMT model of the UK mainland transmission system (using PSCAD). This will enable NESO to perform detailed simulations to identify transients, SSOs, and other phenomena that their current RMS (DIgSILENT) models cannot capture.
New generators are already required to submit a model as part of the GC0141 modification. However, there is an obvious challenge with creating a nationwide model, as much of the existing network lacks EMT models.
With practical engineering experience, it’s possible to estimate or calculate the necessary parameters for EHV cables, overhead lines, transformers, and simple shunt compensation equipment. However, modelling existing generators (and systems like SVC or HVDC) is much more complex, and NESO lacks the resources to manage this. Therefore, NESO is passing this requirement on to existing generators through the GC0168 modification.
For engineers interested in large-scale EMT studies, it’s important to note the significant challenges in performing such simulations, especially on networks with multiple large synchronous machines. While a few countries (like Australia, Canada, and France) have successfully carried out such simulations, this remains a formidable challenge.
Inverter Based Generation
EMT modelling for existing inverter-based generation presents an interesting technical challenge. Once a plant is built, tested, and commissioned, OEMs often move on and no longer provide active support. The plant may have changed hands, and the necessary documentation may no longer be available. At a minimum, re-engaging with the original OEM and testing the equipment will be required to identify its version.
Where things become more complicated is that the original controllers were designed and tested to meet the original Grid Code requirements at the time of construction. The challenge lies in the fact that the OEM models for the inverter and PPC hardware are based on older versions of firmware specific to the installed products. Many OEMs have only recently begun developing EMT (PSCAD) models of their hardware, so these models may not be readily available.
The question then arises: what should the EMT model represent? Should it reflect the actual existing system, for which no OEM model is likely available, or should it be based on the latest version of the OEM firmware/software? If the latter, this would imply that the existing site needs to be updated to meet the latest Grid Code, which may not always be possible due to hardware limitations.
The GC0168 requirements for IBR generation might seem manageable if the IBR is relatively new and the OEM is still in business, with an available EMT (PSCAD) model for the existing generator. However, for older sites (such as wind farms in Scotland), this may not be the case, or the OEM may charge significant fees.
Generic WECC/IEC models could be used, but this undermines the purpose of a specific EMT model, where accuracy is key.
Thermal Based Generation
Thermal generation plants are technically more complex than inverter-based generation, but standard modelling approaches have been established for decades, with typical AVR, OEL, UEL, and PSS models defined in the IEEE 421.5 standard. These models are well-understood in the industry, and PSCAD includes some (though not all) of these in its library. However, they can also be replicated directly in EMT (PSCAD) from the ground up if necessary. One issue with the standard models is that they are designed primarily for small signal stability analysis, which is often overlooked. As a result, they may not fully represent a system undergoing large disturbances.
The more challenging aspect of thermal generation is modelling the governor system. Machine governors are complex, with many local variations. While there is an attempt at standardization in the IEEE TR77 document, it is incomplete. Some OEMs (e.g., GE) adhere fairly closely to standard governor models like GGOV1, but many others do not. These models are not generally available in PSCAD libraries, and developing an accurate governor model from scratch, then testing it, is no simple task.
Summary
The GC0168 process is driven by NESO’s broader goal of creating a nationwide EMT model for the UK mainland transmission system and its main generators. While this is a commendable (albeit ambitious) goal, it faces numerous practical and technical challenges. Some plant owners may be fortunate enough to have a good relationship with their OEM, which may provide an existing EMT (PSCAD) model. However, many will not.
Although GC0168 is not yet approved, it is coming soon, so it’s important to start planning for its implementation. Aurora is expert in PSCAD modelling, so if you’d like to learn more, please get in touch.
