Navigating NESO Grid Code Compliance

Obtaining an ION from NESO is becoming a more and more challenging task. Historically this was driven primarily by construction completion and testing. In more recent years this is being driven by meeting the paperwork and simulation studies requirements. Power system studies and analysis have moved from being a side issue to construction, to a critical path item – if they are not started early enough it is likely that the project will be delayed and possible revenue lost.

In recent seminars and workshops, NESO have advised repeatedly in compliance seminars that the compliance process for studies is becoming increasingly difficult and that an ION will not be given until the requirements are met. NESO reviews are detailed and thorough and the smallest of issues are being picked-up and questioned. We would therefore strongly recommend starting the studies process as early as possible.

With the introduction of GC0141 power system studies for RMS and EMT models are often now critical path items, and failure to complete the studies in time, is one of the most common reasons for project delay. Furthermore, two further changes are expected to the required power system analysis requirements in the short term, which will increase the duration of the systems studies and negatively impact ION dates:

• STCP19-3 SLA – this proposal will increase the review period for studies from 15 business days to 20 business days. Note that this is per submission, so consider that many projects have multiple UDFS submissions.
• Sub Synchronous Oscillation – NESO have released a Guidance Note requiring SSO studies to be carried out. This is not in the Grid Code but often in Appendix F of the BELLA. This study can only be started once the EMT model is complete, and would typically take between 12-22 weeks to complete and sign off.

The following sections detail the typical requirements and issues for each of the main simulation and study case requirements.

RMS Simulations

GC0141 requires (and Planning Code PC.A.9.3) that developers submit an RMS DIgSILENT PowerFactory model of the plant for review by NESO. This model has a number of specific requirements that must be adhered to in order to be accepted. The model must also be provided with a validation report and a user guide.

• The model must be unencrypted and not use any external DLL or reference files.
• The RMS model has to include all the control systems and protection.
• The model report must include an overlay of FAT measurements for Fault Ride-Through testing.
• The model must use a simulation time step of 10 ms.
• The model must be tested against a model IEEE 9-bus network.

The RMS scope is technically not too difficult, but is becoming an increasing logistical problem because the Inverter and PPC contain intellectual property and are often provided by different manufacturers.

Four main approaches are used, and it is essential that the inverter and PPC vendors agree the strategy for this submission at an early stage of the project. At present, Method 2 or 3 are the most common approach.

1. Inverter and PPC OEM agree to provide unencrypted models at the start of the project and these are used all the way through for all grid code simulations. Aurora prepare the validation report and user guide.This is the ideal way, but usually not possible due to intellectual property challenges from the OEMs.
2. Inverter and PPC OEM provide encrypted models as normal for the main studies. At model handover stage, Aurora submit the encrypted model to a trusted 3^rd party (such as DIgSILENT) and the OEMs submit their unencrypted models to the 3^rd party. The 3^rd party swaps out the encrypted bits and submits to NESO. Aurora provide a model guide and validation report. The validation report only works if it is a 1-to-1 swap of encrypted elements with unencrypted elements.  One of the common problems here is that the black box model is designed to run at a lower time step of say 1ms, while the NESO model must run at 10ms.
3. Either one of the Inverter OEM or PPC OEM provides an equivalent generic model (such as WECC) and the other OEM maintains an encrypted model. Aurora then creates a new model based on the open model, and validates the new model and prepares a user guide. Aurora then provides the model to the OEM who still has an encrypted system, and the OEM removes the encrypted system replaces it with an unencrypted version and submits to NESO. This approach is similar to Method 2 above, but a common problem with this approach is that the PPC or Inverter open model does not have enough functionality to meet the NESO requirement, and therefore problems can arise.
4. Both inverter OEM and PPC OEM provide an equivalent generic model (such as WECC) to Aurora. Aurora then undertakes a validation of the WECC models, prepares the validation report and user guide for submission to NESO. This approach facilitates the use of aggregated modelling techniques. This approach is similar to Method 3, but is not favoured as usually the PPC and Inverter open model do not have enough functionality to meet the NESO requirement, and therefore problems can arise.

EMT Simulations

The EMT scope is much more complex, as PSCAD has less standard models and is substantially harder to use. For this scope to work smoothly, the inverter vendor and PPC vendor must supply satisfactory models with good guidance document and be readily available to provide technical support if needed. A common issue with many vendors is that they are relatively new to PSCAD and have limited inhouse expertise, therefore when problems arise they are not able to support and resolve issues quickly. Problems with vendor models are frequently encountered and this can lead to delays, rework and cost overrun – use of reliable and established vendors and consultants helps to overcome this issue.

The simulations in PSCAD are slow and some can take ½ an hour or longer to run. Setting up and configuring the study scenarios is also non-trivial and much harder than the RMS equivalent. Where issues such as oscillation and poor response are encountered with vendor models rerunning studies can take a significant amount of time.

The EMT modelling scope has a number of specific requirements that must be adhered to in order to be accepted. The model must also be provided with a validation report and a user guide.

• The model must work on PSCAD version 5.0.2, Intel FORTRAN v19 and Microsoft OneAPI 2019 (or later versions).
• The EMT model has to include all the control systems and protection.
• The model report must include an overlay of FAT measurements for Fault Ride-Through testing.
• The model must use a simulation time step of 10 µs.
• The model must be tested against a model IEEE 9-bus network.

Factory Acceptance Test of Inverters for Fault Ride Through

NESO have recently confirmed that inverters offered should have been Factory Acceptance Tested (FAT) against UK Grid Code requirements, in section PC.A.9.7.1 & PC.A.9.7.2 for Fault Ride Through (FRT) performance.  This has created number of difficulties as many OEMs have not carried out this requirement. Or have simply carried out generic FAT tests (to other Grid Codes).  

The Client (and NESO) need to be aware that there is inherently going to be differences between the FAT results and the site-wide model. Providing as much info as possible about the FAT configuration will help us get a closer match. There will always be differences between a physical FAT and an RMS and EMT simulation, so these need to be reviewed early to identify any differences and understand the reasons for the discrepancies.

Sub Synchronous Oscillations

Furthermore, NESO are now generally requiring Sub Synchronous Oscillation (SSO) studies on all sites, based on a recent guidance note. It is important to note that the SSO study can only be carried out after EMT modelling is complete.

The SSO study can take up to 2 months to complete and around 4-6 weeks for NESO to review. The requirements for these studies and the consequences if an oscillation is found are not well defined, the most common solution is to try and retune the various control loops that are responsible for the oscillation – this is not a simple or quick task. If the controllers are retuned then all of the Grid Code FRT, FFCI, frequency and voltage studies will need rerunning for both the RMS and EMT studies will need rerunning.

In a best case scenario, where the SSO passes without requiring any intervention, or modification the study requirement will add around 12 weeks to the project. In the scenario where problems are encountered, the retuning of controllers, rerunning the SSO study and updating all the earlier studies   requirement the study could take up to 36 weeks before final approval.

Grid Forming Inverters

Grid Forming Inverters are becoming more common within the industry, and whilst these bring many technical benefits, they also bring a number of challenges and additional study requirements that must be complied with in section PC.A.5.8 and ECP.A.3.9. The technical requirements and challenges of Grid Forming Inverters are therefore expected to significantly delay achievement of receiving an ION.

Summary

The GC0141 requirements for open RMS models and EMT models, validation of inverter FAT results as well as the need for SSO studies have significantly increased the complexity of power system analysis for all Grid Code compliant projects. It should however, be recognised that NESO are facing increasing and mounting pressure of the increasing complexity brought on by higher penetration of renewables on the system, and the associated technical challenges of future networks. This leads to evolving requirements in the Grid Code and increased level of scrutiny on projects.

There are a number of possibilities available to speed up the process, using automation and parallel working, but these are not as simple as would first appear, as the studies are heavily interdependent and generally need to be carried out sequentially. 

From study kick-off we would suggest allowing around 38-66 weeks to gain ION. Solar PV and wind plants would be expected to be at the lower end of this range and larger more complex Storage plants at the higher end.

The following key requirements must be met. 

• Factory Acceptance Test Results for the Inverter for Fault Ride Through and Low Voltage Ride Through events compliant with NESO Grid Code.
• Unencrypted RMS DIgSILENT model that uses a 10ms time step.
• PSCAD model that is compatible with PSCAD v5.0.2, Intel FORTRAN v19, Microsoft OneApi 2019 and a uses a 10µs time step.
• Understand the impact and role of Grid Forming Inverters in system studies.
• Be clear on the process for SSO and the potential delays this can cause.

Typical Flow Chart and Timeline