Mythical 'problem' with IEC 61000-3-2

By J. M. Woodgate B.Sc.(Eng.) C.Eng.  MIET SMIEEE FAES Hon FInstSCE MIOA

Mythical 'problem' with IEC 61000-3-2

A 'problem' keeps surfacing about IEC/EN 61000-3-2, and it's not about meeting its requirements, but that its Class A emission limits are alleged to be inconsistent with the Compatibility Levels in IEC 61000-2-2. 

The (faulty) reasoning goes like this. The compatibility levels are in terms of voltage distortion, but the emission limits are currents. They are related by an impedance. The reference supply impedance given in IEC TR 60725 is 0.4 + j0.25 ?. The TR doesn't say, but the impedance can extended to harmonic frequencies by changing the expression to 0.4 +jh0.25 ?, where h is the harmonic order number. It must be right, because j0.25 is clearly an inductive reactance and that is proportional to frequency.

We can see what happens when we take the compatibility levels from IEC 61000-2-2 and the Class A emission limits from IEC 61000-3-2. There is a complication. The compatibility levels for the Low Voltage (LV) network include stuff coming down from the High and Medium voltage (HV and MV) networks. Only a 'sharing factor' of 25 % of the compatibility level was considered to be available to the loads on the LV network. The result is shown in Figure 1.

Figure 1 Apparent incompatibility; any column taller than 1.0 is a violation

However, current opinion is that this doesn't apply to even harmonics and odd harmonic orders divisible by 3 (triplen harmonics). The sharing factors for these are not universally agreed, but may well be 100%. This is because even harmonic levels on the HV and LV networks are usually very low, and triplen harmonics do not propagate through the type of transformers used in the public networks. Even with this adjustment, the 33rd and 39th harmonics still appear to violate the compatibility levels. But, and it’s a big BUT, the hidden assumption is that there is no other load on the feeder except the one doing the emitting.  This is obviously unrealistic.  We only need to add an additional 10 ohm load (5.29 kW – a small faction of the feeder capacity) to obtain a very different result, as shown in Figure 2.

Figure 2 Realistic comparison of Class A limits and compatibility levels

We can see that there is a reasonable margin between emissions and compatibility levels for every harmonic, and the ratio is particularly low for the important low-order harmonics.  Of course, for more heavily-loaded feeders, the ratios for high-order harmonics are further reduced.

There is a trend, however, for modern techniques to result in higher emission levels of high-order harmonics, so it is necessary to pay attention to them. They are not good news for capacitors for power-factor correction, due to high currents, and for inductors, due to high voltages.