CT saturation from DC offset — X/R ratio and time-to-saturate

The [guide on current transformer protection class (5P/10P) and knee-point voltage](/guides/practical/stroomtransformator-beveiligingsklasse-5p10p-knikpuntspanning) covers saturation that arises when the primary fault current exceeds the CT's accuracy limit factor, and saturation caused by unequal remanent flux between CTs around a differential zone. This article covers a third, separate mechanism: saturation caused by the DC component (DC offset) inherently present in an asymmetrical fault current, right after the fault occurs.

Where the DC offset comes from

When a short circuit occurs on a predominantly inductive system, the fault current does not begin immediately as a pure, symmetrical sine wave. Depending on the point in the voltage cycle at which the fault occurs, the fault current contains an exponentially decaying DC component on top of the symmetrical AC component — the well-known "asymmetrical" waveform of a fault current right after inception, with a peak value considerably higher than the purely symmetrical amplitude. This DC component gradually dies out with a time constant determined by the ratio between the reactance (X) and resistance (R) of the fault path: τ = X / (2πf·R).

Why a higher X/R ratio increases the risk

The higher the X/R ratio of the system at the fault location, the more slowly the DC component decays (a larger time constant τ) and the greater the asymmetry of the fault current. Systems close to a large power transformer or generator typically have a higher X/R ratio than systems further out in the distribution network, and therefore run a greater risk of a prolonged, strong DC offset during a fault.

The effect on the CT: time-to-saturate (ts)

A current transformer that must transform the primary current (including the DC component) to the secondary side effectively integrates the applied flux in its core. The DC component of the fault current contributes disproportionately to this flux build-up compared with the symmetrical AC component, because the DC component's flux does not average out over a cycle the way a pure sine wave does. As a result, a CT that would stay well within its accuracy limit for a symmetrical fault current of the same amplitude can still saturate under an asymmetrical fault current with a substantial DC offset — often within the first few cycles after fault inception. The moment at which this happens is called the time-to-saturate (ts): the higher the X/R ratio and the more heavily the CT is already loaded by the symmetrical component alone, the shorter ts becomes.

Why this can mislead a protection relay

Many protection functions (particularly high-speed differential and distance protection) must make correct decisions precisely during the first cycles after fault inception — exactly the time window in which DC-offset-driven saturation is most likely. A distorted, saturated secondary signal during those first cycles can present a relay with an incorrect amplitude or waveform, with possible consequences ranging from delayed or missed operation to false operation — the latter especially relevant for differential protection, where unequal saturation between the CTs on either side of the protected zone (as also covered in the guide on protection class and knee-point voltage) produces an apparent differential current.

Note: modern protection relays apply algorithms (among others based on harmonic recognition or adaptive filtering) specifically designed to limit the influence of DC-offset-driven CT saturation. This article covers the underlying physical phenomenon, not the full internal compensation technique of any specific relay make.

Practical relevance

When specifying a CT for an application with an expected high X/R ratio (for example close to a large transformer or generator), the expected time-to-saturate under the maximum asymmetrical fault current should be considered in addition to the rated accuracy limit factor (ALF) and knee-point voltage — a CT dimensioned only against the symmetrical component can saturate sooner during a real, asymmetrical fault scenario than the rated protection class would suggest.

Common mistakes

Using only the symmetrical (RMS) fault current when checking CT dimensioning, without accounting for the extra flux contribution of the DC offset at a high X/R ratio.
Assuming a correctly specified protection class (for example 5P20) rules out any form of saturation — the class specification is typically based on the symmetrical accuracy limit current, not automatically on the asymmetrical peak including DC offset.
Failing to include the X/R ratio at the fault location in the system study — the same CT may function without issue at a low X/R ratio and still saturate at a high X/R ratio elsewhere in the same network.
Attributing a false differential operation on a heavy external fault solely to remanent flux differences, without considering the contribution of DC-offset-driven saturation during the first cycles after fault inception as a possible factor.