Page 39 - CE_Industral_Journal_2014

stays at its level, should be a rare situation - normally CTs at a highly increased R for a short time span. Fig. 7 shows
can always be selected in a way to avoid steady-state how the impedance trajectory, for a quite heavily saturated
saturation. One exception could be a weak feeder from current in the transient state, enters from the right and
a bus bar with a small CT fitting the nominal feeder data first reaches an X (mark 1) slightly below the steady-state
- if several strong infeeds to the bus bar feed a fault on fault value but at an R about 10 times as high as the
the weak outgoing feeder then its CT might show steady- steady-state fault value, then (after a time usually too short
state saturation. The problem with this is that, whatever the to actually trip) continues by wandering to even higher R
relay’s interpretation of this corrupted signal shape is, will and about twice the target X until it changes its direction
stay like this for the fault duration. Blocking derived from toward the target (mark 2), traversing in loops that go to
the amount of 2nd harmonic will fail since symmetrical even higher X values. This takes about 180 ms in this
saturation only contains odd harmonic numbers. Much example and, depending on the R and X setting of the trip
more common is transient saturation, present as long as zone, will lead to a trip delay of about the same duration if
the transient DC offset - that is always present when a the upper zone boundary is not far above the steady-state
fault current otherwise would have to ‘jump’ from the pre- X of this fault. For slightly changed data the first low-X
fault to the fault value at fault inception - has not noticeably peak might actually trip an instantaneous zone 1 while the
decayed. Since this offset depends (amongst others) on steady-state fault impedance really is in zone 2, a kind of
the current phase angle at fault inception it will differ in the false trip one might not expect from a current ‘reduced’ by
three phases for a three-phase fault, so the effect on the CT saturation. So for distance relays in transmission as
relay also depends on the implemented or selected inter- well as distribution grids this test is really worthwhile. This
phase treatment (cross-blocking). For two-phase faults example also shows how severe a transient saturation
the effect is the same in both phases since the currents can be although there is no steady-state saturation at all.
are just mirror currents to each other. Now let’s have a
look at some protection principles and possible effects of
CT saturation:

1. Definite-time overcurrent stages: They typically
trigger phase-wise on the peak value of the current. Their
setting is calculated from the steady-state RMS value, i.e.
for non-offset signals. Stages with threshold settings that
are in the magnitude of fault currents that could cause
transient saturation typically are high-set instantaneous
stages. Except for extreme transient saturation, e.g. during
an auto-reclosure and with a CT showing high remanence,
the current will always overcome the set threshold value
before the transient saturation sets in.

Overcurrent relays are typically found in distribution
grids with a short time constant (low L/R) so the effect
of transient saturation lasts for only roughly 100 ms. This
would be the trip time delay to be expected if the relay
function is indeed impaired by strong transient saturation.
Relays that do not evaluate the peak value but e.g. take
the fundamental component are more prone to show trip
delays, but also restricted to the mentioned time span.

2. Distance protection: These relays often calculate
the impedance (as a measure for distance) by analyzing
a sample data window of e.g. one cycle of nominal
frequency and take the fundamental of voltage and current
to calculate the impedance. As tests and theoretical Fig. 7: Impedance trajectory, transient CT saturation
considerations show, transient saturation leads to an
apparent increase in Z magnitude while rotating the Z 3. Differential protection is a special case: Since it is
angle toward the R axis. During the transient phase this well-known that differing saturation at both ends during
might even lead to an X value below the target value but through-fault condition might lead to a false trip, those

INDUSTRY JOURNAL 37