Surge Arrester Testing: Leakage Current and Power Factor on MOV Arresters

Metal oxide varistor (MOV) surge arresters are among the most passive devices in a substation, no moving parts, no scheduled oil changes, no contact inspections. They sit on transformer bushings, bus structures, and cable terminations and perform their function silently until they fail. Because they rarely show visible external signs of degradation before failure, periodic electrical testing is the only reliable way to assess condition.

MOV arresters replaced the older silicon carbide (SiC) gapped arrester design in most applications from the 1980s onward. SiC arresters required a series spark gap to block normal power-frequency voltage from continuously energizing the varistor blocks. MOV zinc oxide blocks have a much more nonlinear voltage-current characteristic, they conduct negligible current at normal operating voltage and clamp voltage sharply at the protection level, so no series gap is needed. The continuous conduction of a small leakage current at normal voltage is both a feature of MOV design and the basis for the primary in-service test method.

How MOV arresters degrade

Every surge event that drives an MOV arrester into conduction causes some incremental degradation of the zinc oxide grain boundaries. Under normal lightning and switching surge duty, this degradation accumulates slowly over decades and the arrester provides reliable service for its rated design life. Two factors accelerate the degradation significantly.

Thermal runaway is the most serious failure mode. If an arrester is applied at a continuous voltage above its maximum continuous operating voltage (MCOV) rating, either because it was incorrectly specified for the system voltage or because sustained overvoltages exist on the system, the leakage current increases, the varistor blocks heat up, the heated blocks conduct more current, and the process accelerates until the arrester fails thermally. The failure is typically violent, with the porcelain housing shattering or the polymer housing rupturing, and it often occurs without prior warning signals during normal operation.

Moisture ingress is the second major degradation mechanism. MOV blocks are sensitive to moisture, and a housing that has cracked, lost its seal at a fitting, or degraded to the point of moisture permeation allows water vapor and liquid water to contact the varistor stack. Moisture dramatically increases the resistive component of leakage current, which produces heat, which accelerates the degradation. In the Southeast, where humidity is persistently high and thermal cycling is significant, moisture ingress is a common cause of premature arrester failure on aging distribution-class arresters.

Leakage current testing

The primary in-service test for MOV surge arresters measures the total leakage current flowing through the arrester at normal power-frequency voltage. Most modern arrester test sets apply a low AC voltage to the arrester (with the arrester disconnected from the line and the test set's high-voltage terminal connected to the arrester's line terminal and the ground terminal connected to the counter/ground lead) and measure the resulting current.

The total leakage current has two components: a capacitive component (leading, 90° ahead of voltage) and a resistive component (in phase with voltage). In a healthy MOV arrester, the capacitive component dominates, the arrester looks primarily like a capacitor at normal operating voltage because the MOV blocks are in their high-resistance, low-conduction region. The resistive component is small. In a degraded arrester, the resistive component grows as the varistor characteristics deteriorate, increasing the total current and shifting the phase angle.

A simple total current measurement is useful for trending but cannot separate the resistive and capacitive components. A measurement that has increased 10–20% over the prior test could indicate resistive degradation, or it could reflect a temperature difference between test dates (MOV leakage current is temperature-dependent, rising with temperature). Separating the resistive component eliminates the temperature ambiguity because the resistive current is the indicator of actual varistor degradation, not temperature variation.

Modern arrester test sets such as the Doble M4100 or Megger DELTA series measure both components using phase-resolved current measurement. The resistive leakage current (also called the watts loss current, W_R) is the diagnostic value. An increase in resistive leakage of more than 50% from the baseline established at acceptance, or a resistive current that exceeds 20–25% of the total leakage current, warrants further investigation. Absolute values depend on the arrester voltage class and manufacturer, a 138 kV station arrester will have different absolute leakage levels than a 15 kV distribution-class unit, so comparison to the arrester's own historical baseline is more meaningful than comparing to a universal threshold.

Power factor testing

Power factor testing on surge arresters applies the same principle as power factor testing on transformer bushings and insulation, the test applies AC voltage and measures the ratio of watts loss to volt-amperes, expressed as a percentage or decimal. A high power factor indicates a high resistive component relative to the capacitive component, which in a surge arrester means varistor degradation.

Doble power factor test sets are commonly used for arrester testing at acceptance and periodic maintenance intervals. The arrester is de-energized, disconnected from the line, and tested in the grounded specimen test (GST) mode with the ground terminal connected to the low-voltage measuring circuit. For station-class arresters with multiple sections in series, each section can be tested independently to localize degradation to a particular part of the stack.

NETA MTS acceptance criteria for surge arresters specify that power factor should not exceed the manufacturer's published value, with investigation required if the measured value exceeds 150% of the factory or prior test value. In the absence of factory data, a power factor above 10% for a polymer-housed distribution arrester or above 5% for a station-class porcelain arrester is a conservative flag for further evaluation.

Insulation resistance testing

Insulation resistance testing of surge arresters with a megohmmeter gives limited diagnostic information on MOV units, the high nonlinearity of the varistor blocks means that a DC megohmmeter reading reflects the high-resistance, off-state condition of the blocks and does not reveal the degradation that manifests at AC voltage. A healthy MOV arrester and a degraded one may both show very high insulation resistance on a megohmmeter.

The one case where megohmmeter testing on arresters is useful is detecting gross moisture ingress, a flooded arrester housing will show dramatically reduced insulation resistance on a DC megger, typically below 1000 MΩ where a healthy arrester would show essentially unmeasurable resistance. The megohmmeter test is a quick screen for gross failure; it is not a substitute for power factor or leakage current measurement for condition trending.

Surge counter and energy monitoring

Many station-class arresters are installed with a surge counter in series with the ground connection. The counter records each discharge event, giving the utility a cumulative record of the arrester's surge duty over its service life. High operation counts, particularly at sites with known lightning exposure or that have experienced switching events, are context for interpreting leakage current trending. An arrester with a high count and rising resistive leakage is a stronger candidate for replacement than one with the same leakage trend but few recorded operations.

Some modern ground leads include a leakage current monitor that provides continuous measurement of the total leakage current without requiring a test set. These monitors, clipped to the ground lead, log current amplitude over time and can provide alert outputs for utilities with SCADA integration. They do not measure the resistive component separately, but a sudden step change in total leakage current is a reliable indicator that something has changed in the arrester's condition.

Testing at acceptance

NETA ATS requires insulation resistance testing and power factor (or leakage current) testing on surge arresters at acceptance. The acceptance test establishes the baseline, the factory-new condition values, against which all future maintenance measurements are compared. Without an acceptance baseline, trending over time is impossible because there is no reference point for "normal" for that specific arrester in that installation. Skipping acceptance testing on surge arresters is common (they are often treated as passive components requiring no commissioning test) and it is a mistake that eliminates the most powerful tool available for detecting gradual degradation.