CT Composite Error and Low-Leakage-Reactance CTs

Short answer

Composite error is the measure that classes P and PR use to say how much a CT is in error with a fault current. IEC 61869-2 defines it as the r.m.s. value of the difference between the primary current and the secondary current multiplied by the rated ratio (3.4.203). Checking it with a real primary current requires enormous sources. For low leakage reactance CTs, the standard accepts an indirect test with the primary open.

Where composite error comes from

Annex 2A of the standard starts from an ideal CT, without saturation. With a linear CT and a sinusoidal primary current, the currents and fluxes are sinusoidal and can be drawn as vectors.

The secondary current Is flows through the winding and the burden, and that determines the induced e.m.f. and the flux. The flux is sustained by the exciting current Ie, which has a magnetizing component and a loss component. The vector sum of Is and Ie is the primary current multiplied by the turns ratio. The difference in magnitude between the two is the ratio error and the difference in angle is the phase displacement (2A.1).

With a small phase displacement, the exciting current splits into a part in phase with the primary current, which gives the ratio error, and a quadrature part, which gives the phase displacement. Ratio error, phase displacement and composite error form a right triangle whose hypotenuse is the composite error (2A.3). The hypotenuse depends on the total secondary impedance, including that of the winding. The split between ratio error and phase displacement depends on the power factors of the burden and of the exciting current.

This triangle only holds under linear conditions. When the core saturates, the exciting current and the secondary current carry harmonics that do not exist in the primary. That is why the standard defines composite error with the instantaneous currents and not as the vector sum of ratio error and phase displacement (2A.4). The standard always treats the primary current as sinusoidal.

Turns correction (when the real turns ratio differs from the rated one) modifies the ratio error but not the phase displacement, and has less influence on composite error than on ratio error (2A.2).

Direct test

The standard method records and digitizes the waveforms of the primary and secondary currents and calculates the composite error by numerical integration according to the definition (2A.5). The annex also describes the traditional methods with analog instruments:

• On a 1:1 ratio CT, the primary and secondary currents are passed through the same ammeter in opposite directions. The resulting current is the exciting current, and its ratio to the primary current is the composite error.
• On a CT of another ratio, a second CT of the same nominal ratio, with negligible composite error, is used together with an ammeter that measures the difference between the two secondary currents.

For classes P and PR, the direct test passes a nearly sinusoidal current equal to the accuracy limit current through the primary, with the rated burden on the secondary (7.2.6.203). The difficulty is practical: with very high primary currents and bar-primary CTs, the distance of the return conductor that will exist in service has to be reproduced.

What low leakage reactance means

The standard defines (3.4.235 and 3.4.236):

• Low leakage reactance CT. One in which measurements made at the secondary terminals, with the primary open, are enough to evaluate its protection performance up to the required accuracy limit.
• High leakage reactance CT. Any CT that does not meet the above. The manufacturer adds a margin to cover the additional leakage flux.

Annex 2C, which is normative, requires four things to be demonstrated:

1. The core is a substantially continuous ring, with uniformly distributed air gaps, if any.
2. The secondary winding is uniformly distributed.
3. The primary conductor is symmetrical with respect to rotation.
4. The influence of the conductors of the neighboring phase and of other phases outside the CT housing is negligible.

If this cannot be demonstrated by drawings to the satisfaction of manufacturer and buyer, tests are compared:

• For the classes TPX, TPY and TPZ, the construction factor Fc is determined, which reflects the differences between the direct and the indirect test (3.4.239). If Fc is below 1.1, the CT is considered of low leakage reactance.
• For the other classes, the composite errors of the whole winding obtained by direct and indirect tests are compared. The low leakage reactance design is considered demonstrated if the direct composite error is below 1.1 times the one derived from the indirect test. The primary current of the direct test is ALF × Ipr for classes P and PR, and Kx × Ipr for PX and PXR.

The standard clarifies that «low leakage reactance» is not a universal property of the CT: it refers to its protection performance, for example to a given class.

Indirect test

With the primary open, the secondary is supplied at rated frequency with a nearly sinusoidal voltage with an r.m.s. value equal to the EALF. The resulting exciting current, expressed as a percentage of ALF × Isr, must not exceed the composite error limit of Table 205 (7.2.6.203, item b). In this method the possible correction of the turns ratio is not considered.

In routine testing, low leakage reactance CTs are tested this way. The others can be tested indirectly by applying a correction factor to the exciting current, obtained from comparing direct and indirect tests on a CT of the same type (7.3.5.203).

Alternative ratio error measurement

Annex 2E, which is informative, describes an indirect measurement of ratio error for low leakage reactance CTs. It gives results very close to those of the direct test and is useful for on-site measurements and for monitoring. The standard clarifies two limits. Routine tests of ratio error are always made directly, because that is the method that gives the most evidence of the low leakage reactance property. And this alternative method never considers the influence of currents flowing near the CT.

What not to do

• Do not use composite error as if it were the sum of ratio error and phase displacement.
• Do not accept an indirect test report without knowing whether the CT is of low leakage reactance or whether it carries a correction factor.
• Do not assume that a CT of low leakage reactance for one class is so for another: the definition depends on the protection performance.

Related topics

Composite error is used in the 5P and 10P classes and in the TP classes. The low leakage reactance CT is verified in the field with the CT Analyzer or with primary and secondary injection. The general view is in current transformer and high voltage. For a first check with classes 5P, 10P or PX, use our free CT burden and saturation tool.

When to call us

At Tiza Group we verify on site whether the CTs in your substation keep the excitation their manufacturer declares and whether they are adequate for your protection. See our instrument transformer testing, in Colombia or elsewhere on request. To review CT saturation in an existing scheme, engineering design and studies, remote and worldwide, includes the analysis. Write to us for a quote.

Related guides

• 5P and 10P Protection CT Classes: Accuracy Limit Factor
• TPX, TPY and TPZ CT Classes: Transient Error
• Current Transformer (CT): Parts, Functions and Tests
• CT Excitation and Knee Point: What It Is and What It Detects
• OMICRON CT Analyzer: Current Transformer Testing

References

• IEC 61869-2, Instrument transformers - Part 2: Additional requirements for current transformers, edition 1.0 (2012), clauses 3.4.203, 3.4.235 to 3.4.239, 7.2.6.203, 7.3.5.203 and Annexes 2A, 2C and 2E.

Frequently asked questions