RIV Measurement and Its Integration with Partial Discharge

RIV measures annoyance, not dielectric severity.

Before the picocoulomb, there was the microvolt. Decades before IEC 60270 standardised partial discharge measurement as apparent charge, manufacturers and utilities were already judging the quality of insulators, line hardware and bushings by listening — literally — to the noise those components injected into the AM radio band. Radio influence voltage is the classic reading of the same physical phenomenon: fast current pulses generated by corona and partial discharge.

The quantity was born as a radio compatibility metric, not as an insulation diagnostic. The original question was whether the equipment interfered with the neighbour's radio, not whether the insulation had an internal defect. That origin explains the three characteristics that still define RIV: the unit in microvolts, which is a voltage induced in the measuring circuit rather than a charge transferred at the defect; the measurement frequency in the broadcast band; and the quasi-peak detector, designed to reproduce the subjective annoyance that repetitive pulses cause a listener on an AM receiver. With apparent charge consolidated from the 1960s onward, RIV moved from principal tool to complementary test, but it stayed mandatory in specific product families and in practically every visual and audible corona test on line hardware and insulators — where the dominant source is external corona, exactly the phenomenon RIV is most sensitive to.

What you will be able to do

• Explain what RIV measures — radio frequency voltage in microvolts, quasi-peak weighted at 1 MHz in the 0.5 to 2 MHz range — and why it relates to partial discharge without ever being equivalent to it.

• Build the RIV measurement circuit per NEMA 107 and the applicable product standards: PD-free coupling capacitor, 300 Ω measuring impedance, blocking filter and a radio noise receiver per CISPR 16-1-1, with documented background verification.

• Calibrate the measurement chain with a traceable RIV calibrator and convert readings between µV and dB(µV).

• Apply the current limits by equipment family — surge arresters, instrument transformers, bushings, breakers and disconnectors, insulators and line hardware — recognising that recent editions moved the RIV requirement to higher voltage classes.

• Plan combined RIV and PD campaigns, interpreting divergence between the two readings as diagnostic information, and document the test to standard.

From the current pulse to the microvolt

Every discharge event — corona at the tip of a fitting, a microdischarge in an internal cavity, a surface discharge on a polluted interface — is a current pulse with a rise time of nanoseconds. By Fourier transform, a pulse of that duration has a spectrum extending from tens of kilohertz to tens or hundreds of megahertz. Each discharge is, in practice, a small wideband transmitter galvanically coupled to the high voltage conductor. The pulse train, modulated by the sinusoidal voltage because discharges concentrate in phase windows, produces radio noise that AM receivers demodulate as audible crackle.

The heart of the receiver is the quasi-peak detector specified in CISPR 16-1-1 for band B, 0.15 to 30 MHz: an envelope detector with a charge time constant of 1 ms, a discharge time constant of 160 ms, and an indicating instrument with a mechanical constant of 160 ms. Those constants are not arbitrary. They were chosen in the 1930s so that the instrument reading would correlate with the subjective annoyance perceived by an AM listener. An isolated pulse charges the detector for 1 ms and then the reading decays; repeated pulses at high rate never let the detector discharge, and the reading rises. The result is a weighting in which the response grows monotonically with repetition rate: for the same pulse amplitude, the quasi-peak reading at a few pulses per second is a small fraction of the reading at hundreds or thousands per second.

Correlation without equivalence

The most common question from anyone arriving at RIV from a PD background is how many microvolts a picocoulomb is worth. The normative answer is categorical: there is no direct conversion between the two quantities. Four independent factors destroy any universal factor.

First, circuit geometry and resonance. The RF voltage that a charge pulse induces across the measuring impedance depends on the stray capacitances and inductances of the complete arrangement — object, busbar, coupling capacitor, cables. The same defect, in the same object, read in two different laboratories, produces different microvolt readings. Apparent charge, calibrated by injection at the terminal, is transferable, which is exactly why IEC 60270 adopted it. Second, spectral content at 1 MHz. The receiver sees only a 9 kHz slice around the tuned frequency, and pulses of the same charge with different rise times — a deep cavity in a winding against corona on a bushing — deposit very different energies in that slice, while attenuation and dispersion along the path to the terminal alter the spectrum that arrives.

The circuit, the calibration and the numbers to memorise

The measurement arrangement standardised by NEMA 107, whose current edition covers measurements from 0.015 to 30 MHz on equipment rated 0.6 kV and above, is topologically identical to the PD circuit, which means practically all the laboratory hardware is reusable. The test source must be a transformer whose own RIV and PD level is demonstrably below the measurement criterion, verified with the object replaced by a discharge-free capacitor. The blocking filter isolates the measuring circuit from conducted mains noise and prevents the object's RF signal from being lost into the low impedance of the source; without it, the background typically makes the measurement impossible. The coupling capacitor is the same PD-free unit from the IEC 60270 circuit, with typical values of hundreds of picofarads to a few nanofarads, closing the RF path from the high voltage terminal to the measuring impedance.

Metrological evidence for RIV follows the same logic as the PD calibrator, with a dedicated generator: the RIV calibrator injects at the object terminal, with the circuit de-energised, a sinusoidal or pulsed signal of known amplitude in microvolts at the measurement frequency, and the receiver reading establishes the scale factor for the whole chain — capacitor, impedance, cables and receiver. Two procedural precautions are non-negotiable and must appear in the report. Background verification: with the complete circuit energised to the measurement voltage but with the object replaced or known good, the background must be recorded, and consolidated practice in product standards requires background no greater than half the acceptance limit — for a 2500 µV limit, background of 1250 µV or less, with reference laboratories working at far larger margins. And the voltage profile: RIV is measured in steps, raising the voltage above the measurement value, typically 10 to 15% above depending on the product standard, holding, and measuring on the way down. This is the same inception and extinction hysteresis logic as PD: discharges that require overvoltage to start remain active at lower voltage, and measuring on the way down captures the conservative condition. Several type tests, surge arresters among them, require the full RIV against voltage curve to be plotted.

Who still requires RIV, and how to read a combined campaign

The RIV map has changed over the last two decades, and outdated training material and many purchase specifications still repeat requirements that current editions have moved or replaced with PD. For breakers and disconnectors, IEC 62271-1 in its current edition applies the type test only from 245 kV upward, where the earlier edition required it from 123 kV — a change made on the basis of positive service experience. The limit remains 2500 µV at 1.1·Ur/√3 into 300 Ω at about 1 MHz. For metal-oxide surge arresters, IEC 60099-4 requires RIV as a type test with the value plotted against voltage, at 2500 µV or less, while routine testing covers internal PD at 10 pC or less at 1.05·Uc. For instrument transformers, IEC 61869-1 keeps RIV as a special test for high classes, historically at 2500 µV at 1.1·Um/√3, but current editions privilege routine PD at 10 pC. For bushings, IEC 60137 applies RIV when specified for radio compatibility, at 2500 µV at 1.1·Um/√3 with measuring impedances from 30 to 600 Ω permitted provided the result is referred to 300 Ω, while routine PD remains the dielectric criterion.

Interpreting the divergences is where the module pays off. High RIV with low PD is the common benign case: an external source, prolific in pulses and poor in charge — corona on a fitting, a point, or a neighbouring object. The correct action is visual inspection, ideally with a UV camera, correction of the arrangement, and repetition. Condemning the insulation on the basis of an externally sourced RIV reading is the classic error of the hurried laboratory. High PD with low RIV is the dangerous case: an internal cavity with few discharges per cycle produces measurable charge on the PD meter, but the quasi-peak detector, penalising low rates, returns modest microvolts. Anyone judging by RIV alone releases equipment with an evolving internal defect. That is the definitive demonstration of why modern routine standards migrated to the picocoulomb: a low RIV does not acquit, whereas a low PD with verified sensitivity does.

Syllabus

• Origin: overhead lines against AM broadcasting, and the first NEMA 107 edition of 1940

• Corona and PD as involuntary radio transmitters: pulse spectra from kilohertz to hundreds of megahertz

• The two observation windows: the wideband low-frequency PD meter against the 9 kHz narrowband receiver

• The quasi-peak detector: 1 ms charge, 160 ms discharge, 160 ms indicator constant, and the resulting repetition-rate weighting

• Why no universal µV to pC conversion factor exists: geometry, spectral content, repetition rate, tuning frequency

• Circuit architecture: PD-free source, blocking filter, coupling capacitor, 300 Ω measuring impedance, receiver

• Calibration, background verification at half the acceptance limit, and the step voltage profile measured on the way down

• Reading in dB(µV): 2500 µV = 68 dB(µV) and 100 µV = 40 dB(µV)

• Current requirements by equipment family and the retreat of the RIV boundary to higher classes

• IEC 60437 as a method standard, not a limit standard

• Combined campaigns: one circuit, two readings, and the two interference scenarios to anticipate

• The decision matrix: high RIV with low PD, and high PD with low RIV

Laboratory work

The practical uses a line insulator or fitting mounted in a test arrangement with a discharge-free source, a coupling capacitor and a combined PD and RIV coupling device, plus a removable artificial corona source in the form of a metal point screwed to the fitting. Stage 1 — dual calibration: calibrate the PD chain with a picocoulomb calibrator per IEC 60270 and the RIV chain with a microvolt calibrator at the tuned frequency; record both scale factors and the background with the source energised and the object replaced, confirming background at or below half the adopted limit. Stage 2 — clean object: take RIV against voltage in steps, raising above the measurement point and reading on the way down, with simultaneous PRPD, and verify the absence of visible corona. Stage 3 — artificial corona: install the point, repeat the profile, and observe RIV jumping orders of magnitude, the PRPD showing the classic corona signature, and the temporal correlation between the two instruments. Stage 4 — verdict: fill in the decision matrix for both measured scenarios and write the report with standard, frequency, impedance, calibrations, background, voltage profile and readings, then defend the conclusion orally.

Key takeaway

RIV never acquits. If RIV is high and PD is low, look outside the object — fittings, points, rings, and a possible broadcast carrier in your tuned window. If PD is high and RIV is low, you have found a low-repetition internal defect, and the picocoulomb is the quantity that decides.

Module 19 of the Atlas Energy Academy · Diagnostic Methods · about 3 hours · Video series (5 × 10–15 min) · Prerequisites: Module 18 — Partial Discharge