Transformer Excitation Current Testing: Reading the Core's Condition

Transformer excitation current testing, also called no-load current testing or magnetizing current testing, measures the AC current drawn by each winding of a transformer when voltage is applied at reduced level with the opposite winding open-circuited. The current drawn is the magnetizing current required to establish the magnetic flux in the core, it is an iron-loss measurement, not a copper-loss measurement. Because the test is applied at a fraction of rated voltage (typically 10% of rated voltage on the Doble power factor test set, or whatever reduced voltage the test instrument applies), the test is safe to perform on de-energized transformers in the field without full high-voltage apparatus.

The distinction from winding resistance testing is fundamental. Winding resistance testing applies a DC current through the winding and measures the resistive voltage drop, it characterizes the copper conductor, the LTC contacts, and the current-carrying path connections. Excitation current testing applies an AC voltage to the winding and measures the resulting current through the magnetic circuit, it characterizes the core steel, the interlaminar insulation, and any parallel magnetic paths created by shorted turns or stray core ground loops. Both tests are run on a de-energized transformer, and both are typically performed in the same field visit, but they cannot substitute for each other.

What the current is measuring

When AC voltage is applied to a transformer winding with the other winding open, the applied voltage must overcome only the winding impedance (which is small at low voltage) and establish the flux in the core. The current divides into two components: the magnetizing component (reactive, 90° lagging), which establishes the flux, and the core loss component (resistive, in phase with the voltage), which represents the eddy current and hysteresis losses in the core laminations. The vector sum of these two components is the excitation current measured by the test instrument. At reduced test voltage, the magnetizing component dominates, the current is largely reactive, and the resistive component representing core loss is relatively small.

The excitation current value is expressed as a percentage of rated winding current, or in milliamperes, depending on the reporting format. For a well-maintained power transformer at 10% rated voltage, the excitation current is typically in the range of 0.1–1.0% of rated current for the winding under test. Large power transformers with low no-load losses (low loss-grade steel) may draw even less. The absolute magnitude is less important than the comparison between phases and between the current test and the factory nameplate value or a prior test record.

Three-phase asymmetry: the normal pattern

On a three-phase transformer with a three-limb (core-form) construction, the most common design for power transformers, the center leg of the core is shared between the flux paths of the two outer legs. The magnetic path length for the center phase winding is shorter than for the outer phases, which means the center phase requires less magnetizing force to establish the same flux and therefore draws a different excitation current than the outer phases. In a three-limb core-form transformer, the center phase (typically B phase) draws more excitation current than the two outer phases (A and C), and the outer two phases should draw similar currents to each other. A typical ratio is A:B:C approximately 70:100:70, where B is higher.

This asymmetry is normal and expected. A test result where all three phases draw the same current would be unusual for a three-limb design and should prompt investigation. For five-limb core designs and shell-form designs, the symmetry pattern is different, factory documentation and design details govern the expected pattern. The key comparison is always current versus the baseline, not current versus a generic threshold. The baseline is the factory acceptance test record that came with the transformer, or the first field test at initial commissioning if factory data is unavailable.

What elevated excitation current indicates

An excitation current that is significantly higher than the baseline on one phase is the primary indicator of shorted turns on that winding. When turns of a winding short together, whether from insulation failure, a through-fault event, or gradual thermal degradation, the shorted turns form a closed loop around the core. This closed loop acts as a short-circuited secondary, loading the core magnetically and requiring additional current to maintain the flux. The additional current shows up as elevated excitation current on the phase with shorted turns.

The sensitivity of the excitation current test to shorted turns is high for turns that are shorted near the voltage end of the winding, where they participate in establishing the full magnetic flux. Turns shorted near the neutral end of a grounded-wye winding contribute less to the magnetic circuit, and the excitation current increase from a small number of shorted turns at the neutral end may be small relative to the baseline variation. For this reason, a normal excitation current result does not rule out all shorted turn scenarios, but an elevated result is strong evidence of the condition.

Elevated excitation current can also result from a stray core ground that creates a circulating current loop around part of the core structure. When a stray core ground creates a conducting loop, the changing flux in the core induces a voltage in that loop and drives a circulating current, which adds to the measured excitation current. The test result looks similar to a shorted-turn result, elevated current on one or more phases, and the core ground test is the procedure that distinguishes the two causes.

What reduced excitation current could indicate

Reduced excitation current, current that is lower than baseline on one phase, is a less common finding and often more ambiguous. A small reduction may be within the measurement variability of the test and does not necessarily indicate a problem. A significant reduction on one phase could indicate that the test connection to that phase has a high-resistance element that is limiting current, a poor lead connection, a tap switch contact with elevated resistance, or an internal open that is not yet a complete break. If winding resistance testing on the same phase shows elevated resistance, the two results point to the same connection problem. If winding resistance is normal but excitation current is low, the geometry of the core or the reluctance of the magnetic path may have changed, SFRA testing can investigate core movement or winding displacement that could alter the excitation current.

Connection and test sequence

The standard excitation current test is performed with the Doble power factor test set or a comparable instrument. For a three-phase transformer, the test is performed one phase at a time (single-phase excitation) with the test voltage applied to the HV terminal of one phase and the other two HV terminals grounded. The LV winding is left open. This configuration energizes one limb of the core predominantly, though the other two limbs see induced voltage as well. Some test protocols also apply three-phase excitation from a three-phase source, measuring the three-phase no-load current pattern simultaneously, this is the most direct simulation of the transformer's operating condition but requires a three-phase variable source.

For delta-connected windings, the test is applied phase-to-phase rather than phase-to-ground, and the interpretation of the three-phase asymmetry pattern requires the transformer's vector diagram. Most modern Doble instruments and equivalent test sets guide the connection sequence and apply the correct test voltage for the winding configuration. Recording the winding temperature at the time of test is important for trending, excitation current has a small but measurable temperature dependence, and comparing results taken at significantly different oil temperatures without correction introduces a systematic error in trending.

Relationship to SFRA and the complete diagnostic picture

Excitation current and sweep frequency response analysis (SFRA) are complementary tests for core and winding integrity, not alternatives. SFRA is most sensitive to physical displacement of windings, mechanical deformation of the winding geometry, and changes in the capacitance and inductance distribution along the winding length. Excitation current is most sensitive to shorted turns and core magnetic circuit changes. A transformer that has experienced a through-fault event should have both tests performed, SFRA to detect physical winding movement, and excitation current to detect shorted turns. The tests are not redundant; each catches conditions the other may miss.

For current transformer excitation testing, the concept is related but the diagnostic purpose differs, CT excitation testing establishes the knee point of the saturation curve, which governs the CT's ability to accurately reproduce fault current for protective relaying. Power transformer excitation current testing at reduced voltage is not measuring the same parameter as CT knee point testing, though both tests apply voltage to a winding and measure the resulting current.