Ratio, polarity and magnetics
A current transformer excitation test energizes the secondary winding with a rising alternating voltage while the primary is open, and plots the exciting current that results. The curve shows where the core begins to saturate, which tells protection engineers whether a relaying current transformer will reproduce fault current faithfully.
A current transformer excitation test, often called a saturation test or a knee-point test, applies a variable alternating voltage to the secondary winding of a current transformer with the primary left open, and measures the exciting current drawn at each voltage step. Plotted together, the voltage and current points form the excitation curve of the core. The curve has a gently rising lower portion, a bend known as the knee, and a steep upper portion where a small increase in voltage demands a very large current because the iron is saturated. The test is called for on relaying-class current transformers, including bushing-type units inside transformers and breakers, window and bar types in switchgear, and multi-ratio units where several taps are available. When a unit carries both a metering core and a relaying core, the relaying core is the one that gets this test. It is a field check of the core itself, independent of the burden connected to it, and it complements the ratio, polarity and burden measurements made on the same device.
A protective relay depends on the current transformer to deliver a faithful copy of fault current, and it must do so just when the primary current is largest and the core is under the most stress. If the core saturates early, the secondary waveform is clipped, the relay sees less current than really flows, and a trip can be delayed or missed. Differential schemes are especially sensitive, because unequal saturation of two current transformers creates a false operating current. Problems the excitation curve exposes include shorted turns in the secondary winding, a core damaged or contaminated during manufacture, an air gap or joint problem, residual magnetism left by a dc test, wrong tap selection on a multi-ratio unit, and a unit that is simply a different design from the one specified. Finding these at commissioning is far cheaper than discovering them during a system fault, when the relay misoperation can turn a local event into a wide outage and cause equipment damage.
A current transformer works because the primary ampere-turns are balanced by secondary ampere-turns, apart from a small part that magnetizes the core. That magnetizing part is the exciting current. In the lower region of the curve the core iron is easy to magnetize, so a small current is enough and the voltage rises almost in proportion. As the flux density approaches the material limit, the core can no longer take more flux, the permeability collapses, and the exciting current climbs rapidly for very little extra voltage. The bend between the two regions is the knee point, and standards define it by a particular slope of the curve on logarithmic axes. The voltage at the knee is a rough measure of how much secondary voltage the core can support before it distorts the output, which is exactly the quantity a protection engineer needs when comparing it against the burden and the fault current. The curve shifts with the number of turns in service, so each tap of a multi-ratio unit has its own curve. Because the iron remembers earlier magnetization, residual flux from a previous dc measurement or a through-current can pull the curve toward earlier saturation, and the unit is demagnetized before and after the test.
The circuit is de-energized, isolated and grounded under the owner's safety procedures, and the primary is left open. The secondary leads are disconnected from the relays and meters, and any shorting links are opened, so that only the current transformer winding is tested. Because the test voltage is applied on the secondary, the primary side can develop a high voltage, and test personnel keep clear of the primary terminals and observe the same precautions as with any energized equipment. A variable alternating source with a suitable rating, such as a variable transformer and a true-rms voltmeter and ammeter, or a purpose-built current transformer analyzer, is connected to the full secondary winding or to the tap being tested. The voltage is raised in steps while the exciting current is recorded, and the sweep is stopped as soon as the knee has been clearly passed so that the winding and the instrument are not overstressed. The voltage is then brought back smoothly to zero, never switched off abruptly, to avoid leaving residual flux. A demagnetizing routine is performed after the test, either with the analyzer's built-in function or by slowly cycling the voltage down. Each phase is tested, each tap that will be used is tested, and the instrument must be in calibration, with its frequency matching the system.
The record shows the voltage and exciting current points that make up the curve for each phase and each tap measured, the secondary winding or core that was tested, and the voltage at which the knee was identified. It also notes the instrument and its calibration status, the source frequency, the date, and the ambient conditions. The as-found state is described as well, for example whether the unit was freshly installed, whether it had been subject to a dc test beforehand, and whether demagnetization was carried out afterward. Having the full set of points and not only the knee figure allows a later reviewer to overlay curves and to see where, and how sharply, a change has occurred.
There is no single pass or fail number here, and this page gives none. The principal comparison is between the measured curve and the typical curve published by the manufacturer for that unit and ratio. The measured knee voltage is expected to be consistent with that curve and with the unit's accuracy class, and the shape should be smooth, without abrupt steps. Curves for the three phases of the same installation, with the same ratio and design, are compared with each other, since an outlier suggests shorted turns or a damaged core. A comparison against earlier results on the same device shows whether the characteristic has drifted. The calculated knee must also be adequate for the burden and the fault current of the protection scheme, which is a design question answered by the relay engineer. The method of defining the knee point and the accuracy classes are covered in IEEE C57.13-2016 and IEEE C57.13.1-2017, while ANSI/NETA ATS-2025 and ANSI/NETA MTS-2023 call for the excitation test on relaying applications; consult your licensed copies for the governing requirements.
The excitation test is part of the acceptance sequence for current transformers used in relaying, and it is repeated in maintenance work when the protection scheme is under review or when a previous result was questionable. Metering-class cores are generally not tested this way. On a unit with multiple secondary windings, each relaying core is tested on its own with the others open or shorted as the procedure directs. Bushing current transformers inside a transformer or breaker are tested from the terminal block, with the equipment isolated. Multi-ratio units are tested on the tap that is in service and, ideally, on the highest tap as well, because the curve depends on turns. Gapped cores and high-accuracy units have a much more gradual knee and need judgment, since the standard knee definition suits only ungapped cores. In every case the manufacturer's published curve is the main yardstick.
The equipment classes and standard sections that call for this test, and the UltraDb forms that carry it.
It is the point on the excitation curve where the core starts to saturate, so a small further rise in voltage needs a large rise in exciting current. Standards define it using the slope of the curve on logarithmic axes. It indicates how much secondary voltage the core can support before the output is distorted.
With the primary open, no primary current is balancing the secondary ampere-turns, so everything the source supplies goes into magnetizing the core. That makes the measured current the true exciting current. The open primary can develop a hazardous voltage, so it is treated as energized.
Raising the voltage into saturation, or applying a dc test such as a resistance measurement, can leave residual flux in the iron. That flux moves the operating point and can cause early saturation during a later fault. Demagnetizing restores a neutral starting state before the unit returns to service.
Normally it is not, because metering cores are designed to be accurate at normal load and to saturate deliberately at modest overcurrent, which protects the meter. The excitation test is called for on cores used for relaying, where behavior at high fault current is the concern.
UltraDb references ASTM, ANSI, IEEE, NETA, NFPA, NEMA, FAA, TIA, and IEC standards descriptively, to identify the published methods its forms are built from. UltraDb and Verlecta are not affiliated with, endorsed by, certified by, or licensed by NETA, ASTM, ANSI, IEEE, NFPA, NEMA, FAA, TIA, IEC, or Megger. Standard names and section numbers are the property of their respective organizations. Always test to the edition your contract specifies. Where a standard is sold through the ANSI webstore, Verlecta may earn a commission on a purchase made through our link, at no extra cost to you. A commission never decides which standards we cite or how a form is built.