17-08-2012, 11:30 AM
Power Quality Application Guide
Power Quality Application Guide.pdf (Size: 296.99 KB / Downloads: 63)
Passive Filters
In Section 3.1.2 of this Guide it was explained why reactive power should be compensated and how this is
best accomplished. Fundamental reactive power is always an onerous oscillation of energy. When
considering harmonic currents it is not so clear that they can be addressed as a second type of reactive
power. Harmonic currents may originate from systems in which there is no energy and where the sign of
the (composite) current matches that of the voltage throughout the cycle (e.g. a phase-angle controller for
an incandescent lamp). The term ‘wattless current’ is sometimes applied to the harmonic current when
there are no substantial voltage harmonics of same orders to multiply them with – the product of current
and voltage for an individual order being zero. However, harmonic currents have a lot in common with
reactive currents:
They are both undesirable in that they require part of the capacity of generators, cables, and
transformers, while contributing nothing to the generation and transport of electrical energy.
They both cause additional losses – because the voltage drop is phase related to the current so the
product is real and non-zero.
Harmonics originate mostly from the power-consuming load and flow back to the energy source,
against the normal energy flow (Figure 1). (An exception is a renewable energy source connected to
the grid by a power electronic converter, where the harmonics flow from the source.) Fundamental
reactive power does not have a defined direction - intake of inductive reactive power is synonymous
with output of capacitive reactive power and vice versa.
Reactive compensation
Reactive current compensators are affected by harmonics (as explained in Section 3.1.2 of this Guide) and
it is recommended that power factor correction capacitors are de-tuned. In fact, some electricity suppliers
require de-tuning.
‘De-tuning’ means connecting a reactor in series with the PFC capacitor so that the capacitor/inductor
combination behaves as a capacitor at the fundamental supply frequency but has a defined behaviour for
harmonic frequencies.
A simple (non de-tuned) PFC is actually part of an acceptor circuit formed with inductive components in
the network, especially with stray inductance from transformers. Resonance will lead to excessive
harmonic currents and to excessive voltage drops in the proximity of transformers affected.
It has been explained that, at the tuned frequency, the magnitude of the voltage drops across the inductive
and capacitive elements is the same but with a 180o phase difference, giving a resultant ‘zero’ voltage drop.
However, at or near resonance, the voltage drop across each element is much higher than that would be
expected across, for example, the network impedance at the point of common coupling. So, considering
the elements individually, each has a high magnitude voltage drop across it even though the resultant
voltage drop across the combination is small. This explains why ‘accidental’ acceptor circuits (e.g a PFC
capacitor with stray inductance) are a problem – the installation is across the capacitive element and sees
these amplified voltages. When the inductive element is added intentionally, the installation is across the
resultant acceptor voltage drop. The excess voltages remain inside the compensator cubicle, say across the
capacitors designed for these voltage values, but at its outside terminals no resonance or magnified
voltages can appear.
It is worth remembering that, especially where single phase non-linear loads are in use, there are harmonic
frequencies at 100 Hz intervals from 50 Hz to well over 1 KHz so there is ample scope for resonances to be
excited.
Combined compensation and filtering
In practice, the functions of reactive power compensation and harmonic current filtering are often
combined. It is usual to set the resonant frequency of the LC circuit at a non-harmonic frequency because
compensators may easily be overloaded. The rating of the reactors is normally given as a percentage of the
rated reactive power of the capacitors at 50 Hz. For example, a 5% de-tuning rate means that 1/20 of the
voltage drops across L and 21/20 drop across C, subtracting to 100% overall. At 20 times the frequency, say
1,000 Hz, the ratio would be reversed, so the resonant frequency where XL and XC are equal lies in the
middle between these two frequencies, to be precise at:
Wattless current
As has already been mentioned, where reactive power occurs in a distribution system (usually inductive
reactive power), part of the energy in the line is in effect not transferred from source to load. Rather it
oscillates from a capacitance to a reactance and back again at a frequency of 100 Hz. For certain intervals
of time voltage and current have opposite polarities (Figure 2). Looking at harmonics, the picture appears
very similar. In Figure 3 the power of the third current harmonic has been plotted in isolation. The power
transferred is the product of the third current harmonic times the voltage present in the line, assuming the
line voltage is still a pure sine wave. It can be shown that the areas above and below the abscissa cancel,
meaning that at average no energy is transmitted. The third harmonic current is therefore absolutely
wattless.
But since harmonics do cause additional losses, there must be some active power associated with them.
This apparent contradiction originates from the incorrect assumption that the supply voltage was free of
any harmonics. This is impossible, since the moment there is any 150 Hz current flowing, it will cause
some – active and probably also reactive – 150 Hz voltage drop. This means that as soon as there is any
additional frequency contained within the current, there will also be a certain amount of the same
frequency in the voltage. Only when both voltage and current of the same frequency are present can active
power occur at this frequency. It should be clear by this point that this will always be the case to some
extent. The resistance in the installation circuit causes voltage drop that is exactly in phase with the current
and therefore results in real power dissipation whether the current is real, reactive or harmonic.
Sample measurements
Fluorescent lamps are the only common device
where putting the most efficient way of
compensation, at the point of origin within the
luminaire, is common practice. This is most
efficient because only real current flows in the
installation wiring, the reactive component
having been compensated within the fitting.
When centrally installed units are used,
combining the reactive current compensator
with the harmonic filter solves several problems
at the same time with the same device. The
advantage of a centrally installed unit, with
appropriate control is that, since not all
equipment operates simultaneously, it is often
possible to install rather less total compensation
capacity than would be the case if all equipment
were locally compensated. It also reduces the
risk of overcompensating motors. Using a
combined filter/compensation device