Redwing Technologies






Jeremy Hayward Pr.Eng



A great deal has been (and no doubt will continue to be) discussed and debated about the minimum Quality of Supply Standards outlined by the NRS-048 document, with the "debate" surrounding the associated issues being labelled by interested parties as everything from constructive to scandalous, with even accusations of inflammatory rhetoric.

However, this article intends to highlight specific consumer problems which can, and have occurred with existing capacitor banks and harmonic filters due to increased "external" harmonic levels on the utility supply.

Consumer PFC design - the basics

All Power Factor Correction (PFC) installations installed by consumers are 'site specific', whether as plain capacitor banks or harmonic filter arrangements. The PFC is generally designed to provide sufficient power factor correction (i.e. VAr's) in the consumers system based on the individual requirements, e.g. for tariff purposes, demand reduction, voltage improvement, etc.

The PFC design must take into account all aspects of the electrical supply system to which it will be connected, in order to ensure no adverse system interaction. Critical aspects that must considered during the design stage include:

  • consumer loads to be operated in parallel with the PFC, with particular emphasis on any harmonic sources such as thyristor controlled loads (e.g. mine winders);
  • existence of any consumer devices or loads which may be affected by PFC operation, e.g. ripple control, certain domestic earth leakage relays, variable speed drives sensitive to supply voltage disturbances;
  • details of the utility supply arrangement to the consumer including fault levels, supply transformer details, utility network impedances, other consumers supplied from the same utility network, etc.

An optimised PFC design should also take into account possible changes to the consumer and utility networks. These would typically include opening/closing of buscouplers, parallelling of supply transformers, etc.. Furthermore, where the possibility exists for future additional installation of harmonic loads, this must also be taken into account.

An additional, and very important, aspect which must be considered are harmonic sources present on the utility's supply network. These harmonic levels are generally increasing due to increased non-linear loads installed by the utility consumers, often without adequate harmonic filtering.

It is therefore essential that the PFC design not only consider prevailing high voltage harmonic distortion in the PFC equipment parameter ratings, but also potential maximum acceptable supply harmonic distortion which could be present in the future. Although the former can fairly easily be determined, without the privilege of hindsight the PFC designer has to rely on existing accepted harmonic limits at the utility high voltage supply for the latter.

In these days of reducing costs wherever possible, it is often not economically viable or in fact technically justifiable to design PFC installations for "external" harmonic levels which significantly exceed the accepted 'limits of the day'.

Revisions to harmonic voltage limits

In a period of about seven years recommended and generally accepted limits for voltage waveform distortion on South African Supply networks has changed twice. In 1992 the Electric Power Coordinating Committee (EPCC) outlined revised guidelines to existing generally accepted harmonic limits, and in 1997 'minimum quality of supply standards' were outlined in the NRS-048 document based on work initiated by the National Electricity Regulator.

A summary of these harmonic limit revisions for significant harmonics generally present on utility supply networks are listed below:

Redwing Technologies

Redwing Technologies

One of the problems with increasing limits, is that the prevailing harmonic levels also seem to follow the trend.

Increased limits for "external" harmonic voltages and existing consumer PFC

1.       Consumers supplied via a utility HV grid

It can be seen although the overall voltage THD limit has remained at 3%, the EPCC and NRS limit for individual 3rd and 5th harmonic voltage have effectively doubled, compared to the pre-1992 limit of 1%.

This increase in harmonic limits can allow harmonic levels on utility supply networks to increase beyond limits upon which consumer equipment was originally designed, thereby shifting responsibility for any related adverse effects in the consumers system onto the consumer.

Case Study -

The consumer had PFC capacitor banks designed and installed at his dedicated 22 kV supply from the utility in 1988, when the prevailing 5th harmonic voltage at the utility 132 kV supply was about 0,5%, or half the recommended 'limit of the day'. These banks operated successfully for over 6 years, after which spurious tripping occurred due to harmonic overload.

Upon investigation, it was found that the banks were tripping due to series resonance amplification of an "external" 5th harmonic voltage source of about 1,5% present on the 132 kV supply, i.e. 50% higher that the accepted recommended limit at the design stage, and three times the original level recorded.

However, as the 5th harmonic voltage limit at 132 kV had in the interim effectively doubled from 1% to 2,1% as listed by the EPCC, the new 5th harmonic voltage was now within limits.

In order to continue to use his PFC, the customer was forced to convert his capacitor banks to harmonic filters at significant cost.

2.       Consumers supplied via a utility MV grid

The voltage THD limit increased from 3% to 5% in 1992 (EPCC limit), thereafter increasing to 8% in 1997 (NRS limit). This represents an effective 270% increase in THD limit.

The situation for individual harmonic voltages is even worse, with an effective 500% and 600% increase in recommended limit for 3rd and 5th harmonic voltage respectively.

Although increased limits alone will obviously not result in consumer PFC problems, it appears inevitable that increasing harmonic limits eventually results in higher harmonic levels on the utility supply networks.

For consumers with existing PFC installed and a dedicated supply from the utility via a utility MV network (typically from 11 kV to 44 kV), the potential for spurious PFC tripping is therefore even greater. Of course if PFC protection is not set up correctly, indications of a problem could first appear through PFC equipment failure such as capacitor can failure.

For example, consider a 'typical' consumer who is provided with a 6,6 kV supply by the utility via a 44/6,6 kV transformer. The consumer had a 4 MVAr capacitor bank installed in 1990 to improve power factor at his 6,6 kV supply, the PFC design being based at the time on existing and future harmonic generation in the consumer's plant, together with a potential "external" 5th harmonic voltage source at the utilities' 44 kV supply of 1% - being the accepted limit. However, by the time 1997 arrived, two sets of revised harmonic limits had allowed the utility to move the harmonic goalposts.

With an increase in MV 5th harmonic voltage to 4% (1992 EPCC limit), the consumer would experience a 5th harmonic voltage at his 6,6 kV supply of about 7% due to series resonance effects. Although not an ideal situation, it would be a liveable for the consumer.

However, the utility is now in a position to 'allow' the 5th harmonic voltage source at their 44 kV supply to approach 6% and still remain within the NRS limit. This consumer would now experience a 5th harmonic voltage at his supply of over 10%, leading to tripping of his PFC and related financial penalties.

The future ?

As small and medium consumers cannot go on strike and presently do not have the option of an alternative utility to stimulate competition (although one wonders whether this will ever occur in SA), they effectively have very little say in respect of imposed limits. Certainly in respect of the increased "external" harmonic levels allowed on the utility supply by revised limits, it is the consumer who will bear the brunt of implementing remedial measures to alleviate any harmonic overloading of PFC. This PFC being correctly designed according to previous harmonic limits.

However, hopefully the latest harmonic limits as specified in NRS-048, although termed both 'constructive' and 'scandalous', are here to stay for at least the foreseeable future. This will provide a certain clarity in what seems over the last few years to have been a grey area in respect of allowed harmonic limits, as well as providing the PFC design engineer with fixed performance guidelines.

Finally, although it is difficult to see how the harmonic limits can be increased beyond NRS-048, I suppose anything is possible. As Winston Churchill once remarked "we must beware of needless innovation, especially when guided by logic".


Top of Page