Learnings

Outcomes

1. Two technical reports on 1) Modelling of North Scotland test system and 2) Modelling adequacy and analysis of SSO were delivered.  

2. Algorithms and Matlab scripts for estimating IBR models and analysing root cause of poorly damped SSO were shared. A user interface was created and demonstrated to NESO. 

3. A PSCAD model of a reduced equivalent North Scotland test case was developed with generic IBR models and network data and operating condition provided by NESO. 

4. Two research publications on In-situ estimation of IBR models for SSO analysis; and RMS vs EMT trade-offs for SSO studies (https://ieeexplore.ieee.org/document/11082650) resulted from this project. 

Lessons Learnt

Although use of pulse response in conjunction with eigensystem realisation algorithm (ERA) avoids tedious dynamic frequency scans and enables in-situ estimation of IBR models, it is challenging to use probing-based methods in a real system. Hence, ‘probing-free’ or ‘offline probing’ methods needs to be investigated and a suitable trade-off between zero (or less) probing and compromise in accuracy needs to be analysed. 
 
In addition, theoretically guaranteed match between estimated and actual model at selected frequencies needs to be established. That way, one-shot multi-sine probing with sufficiently high frequency resolution will guarantee very close match to improve the accuracy of the estimated IBR models and hence, that of SSO analysis. 
 
For the North Scotland test system, IBR plants need to be modelled properly with power plant controller (PPC) and voltage regulation device (STATCOM) at the point of interconnection. Such IBR plant models should be sanity checked by benchmarking the admittance spectra with respect to that of actual IBR plant models.