Earlier, we shared a Technical Insight piece regarding the Stage 3, ENA G5/5 harmonic study for an Industrial Plant in the UK, Part 1 – Simulation and Analysis by Kerim Ozer, BSc, our Principal Power System Consultant and in-house power quality expert. We are now happy to share the follow-on, Part 2 – Harmonic Filter Selection and Design.
To recap, Enspec was commissioned to conduct a Stage 3, ENA G5/5 harmonic study for an Industrial Plant in the UK, which includes a total capacity of >5MW induction machine drives – a mixture of 6 and 12 pulse thyristor rectifiers. The site is to be connected to the 33kV grid via a 20MVA, 33/11kV grid transformer. The drives are located at the 415V side of the various 11/0.415kV transformers.
The purpose of this study was to investigate the impact of the site in terms of harmonic distortions at the 33kV Point of Common Coupling (PCC) using DIgSILENT PowerFactory.
The upstream 33kV network was modelled by an external grid and a frequency-dependent impedance. The range of equivalent Thevenin system impedances for each harmonic order from h=100Hz to h=5000Hz was provided in impedance loci form by the Distribution Network Operator (DNO).
A script developed in DIgSILENT was used to analyse the harmonic voltages across the entire impedance loci. The resolution used for the impedance loci was approximately 5000 nodes for each impedance locus.
The analysis found that the site is not compliant with the G5/5 Stage 3 limits due to the harmonic limit exceedances at various harmonic orders – 5th, 7th, 11th, and 17th orders. It was also found that the site did not result in a considerable impedance escalation between the 2nd and 100th harmonic orders. Thus, the exceedances were caused by the incremental increase (Vhc), not background harmonic amplification.
Based on this analysis, the site required harmonic mitigation measures in line with ENA EREC G5/5. To mitigate the high harmonics, it is recommended that either an appropriately sized passive harmonic filter or multiple active harmonic filters should be installed. Regarding the cost-effective perspective, a single passive filter was selected and designed for connection at the 11kV switchboard.
Considering the G5/5 compliance and reactive power requirement of the site, the performances of various harmonic filters were investigated based on different Quality Factors (0.125 – 1.0).
In conclusion, the installation of an 11kV, 4.0 MVAr C-type filter with the appropriate tuning frequency (275Hz) and quality factor (1.0) has been proposed to the client.
The studied worst-case harmonic voltage distortion (THD) results below clearly indicate that the C-type filter with the Quality Factor of 1.0 ensures optimum performance against the G5/5 harmonic limits at the 33kV PCC – the THD could be mitigated from 5.50% to 1.60%.
