Can a power quality event be proven to have caused equipment damage?

Sometimes, and the honest answer in many matters is no. Proving causation requires three things to line up: a recorded event of sufficient severity, an equipment immunity specification the event actually exceeded, and a physical failure mode consistent with that specific electrical stress. Where all three align, the opinion is strong. Where the event is inside the ITI or SEMI F47 ride-through envelope, or the failure mode indicates a thermal or mechanical progression rather than an electrical transient, the attribution fails — and it is better to know that before an expert designation than after.

Who is responsible for harmonic distortion at the point of common coupling?

IEEE 519 allocates it in two directions. The customer is responsible for limiting the harmonic current injected into the utility system, with limits scaled by the ratio of short-circuit capacity to load current. The utility is responsible for maintaining voltage distortion at the point of common coupling within limits, given customers who meet their current limits. Most harmonic disputes turn on measurement location and on whether the load actually present matches the load the allocation was based on. Both are answerable from data if the monitoring was set up correctly.

What is the difference between a sag, an interruption, a transient, and a surge?

IEEE 1159 defines them, and the distinction matters because the damage mechanisms differ. A sag is a short-duration reduction in RMS voltage, typically 10 to 90 percent of nominal for half a cycle to a minute — the most common cause of production loss and the least likely cause of physical damage. An interruption is a complete loss of voltage. A transient is a sub-cycle event, either impulsive such as a lightning surge or oscillatory such as a capacitor switching event. Loose use of the word 'surge' to describe all four is the most common source of confusion in power quality claims.

Does a power quality monitor recording an event prove the utility was at fault?

No. A monitor records what happened at the monitor's location; it says nothing on its own about where the disturbance originated. Sags recorded at a customer facility are frequently caused by on-site events — large motor starting, an internal fault, capacitor switching, or a downstream breaker operation — and the recorded waveform often distinguishes an upstream from a downstream source. Determining direction requires the waveform capture and the facility configuration, not the summary event table.

How long is power quality data retained, and what should be preserved?

Recorder buffers vary from days to a few months depending on configuration and trigger frequency, and utility event records follow their own retention schedules that are typically not long. Preserve the raw waveform captures rather than the summary report, along with the monitor's configuration and calibration records, and request the utility's feeder event history in writing early. Summary reports discard the waveform detail that makes attribution possible, so a case built on a summary table alone is usually a case that cannot be finished.

Power Quality Expert Witness | Jay Prigmore, Ph.D., P.E.

Attribution is the real question

Most power quality engagements are not really about characterizing the disturbance. Modern meters do that well. They are about attribution — whether a documented event caused the failure being claimed, and whether responsibility sits with the utility, the facility, a neighboring customer, or the equipment itself.

That question is answered with waveform records, equipment withstand capability, and physical failure evidence together. An event that coincides in time with a failure is a starting point, not a conclusion; a voltage sag well within a device's ride-through capability did not damage it, whatever the timestamps suggest.

Phenomena investigated

Harmonics — distortion measurement and modeling, resonance with power factor correction capacitors, transformer and neutral conductor heating, and evaluation against IEEE 519 limits at the correct point of common coupling.
Voltage sags and swells — characterization, source direction, and equipment ride-through assessed against the ITIC/CBEMA curve and SEMI F47 where process equipment is involved.
Transients and surges — switching transients, capacitor bank energization, lightning-induced surges, and surge protective device performance and coordination.
Voltage unbalance — and the disproportionate motor heating it produces, which is a frequent cause of premature motor failure.
Flicker — arc furnace, welder, and large motor starting effects.
Notching, noise, and grounding problems — including the ground loops and bonding defects that get misdiagnosed as utility-side power quality issues.

Common dispute patterns

Equipment damage claims against a utility — whether the delivered voltage left the applicable service standard, and whether the failure is consistent with that excursion.
Harmonic responsibility — allocating distortion between a facility's own nonlinear load and the upstream system.
Process interruption — whether a sag that tripped a line should have been ridden through, and whether the equipment met the specification it was bought under.
Capacitor and resonance failures — where added correction created a resonance nobody modeled.
Repeat failures with no obvious cause — where power quality is suspected but has never actually been measured.

Measurement and analysis

Where monitoring is needed, PEFG specifies the instrumentation and monitoring points, then interprets the results — which is normally where these engagements succeed or fail. Power quality monitoring produces very large volumes of data, most of it irrelevant to the question, and a monitor placed at the wrong point answers a question nobody asked.

Existing records are often more valuable than new monitoring: utility meter and relay event captures, drive and UPS fault logs, and facility monitoring data frequently already contain the event, and they have the considerable advantage of being contemporaneous rather than recreated after the dispute began.

Related free tools

PEFG publishes free power system calculators at powerengcalc.com, including voltage drop, motor starting, and transformer thermal life. Those are for engineering estimation; a power quality investigation requires measurement and analysis of the actual system.

Standards and reference curves applied

IEEE 1159, Recommended Practice for Monitoring Electric Power Quality — the classification and terminology that keeps an argument from collapsing into imprecise language about "surges".
IEEE 1453, Standard for Measurement and Limits of Voltage Fluctuations and Associated Light Flicker on AC Power Systems — the P_st/P_lt flicker indices and the emission limits a flicker complaint is measured against.
IEEE 519, Recommended Practice and Requirements for Harmonic Control — the allocation of harmonic responsibility between the utility and the customer at the point of common coupling.
IEC 61000-4-30 — power quality measurement methods and instrument classes, which determine whether a recorded event can be relied on at all.
The ITI (CBEMA) curve and SEMI F47 — the voltage sag ride-through equipment is expected to tolerate, and the benchmark a susceptibility claim is measured against.
ANSI C84.1 — steady-state voltage ranges, separating a genuine service voltage problem from normal utilization variation.
IEEE C62.41 and the C62 surge series — expected surge environment and the performance of surge protective devices.

A worked example

Illustrative — attribution, not detection

A plant records a 62% voltage sag lasting 4.5 cycles. Two hours later a drive fails. The claim asserts the sag destroyed the drive.

Detection is the easy half; the monitor already did it. Attribution is the dispute. A 62% sag for 4.5 cycles sits in a specific region of the ITI curve and of SEMI F47, and equipment meeting either is expected to ride it through. If the drive was specified to ride through it and did not, the more probable explanation is that the drive did not meet its own specification — which is a product claim against the manufacturer, not a service claim against the utility.

The two-hour gap is the other problem. Electrical damage from a sag is essentially immediate: undervoltage trip, DC bus collapse, or component failure occurs during or within cycles of the event. A failure two hours later is more consistent with a thermal or mechanical progression that was already underway, and a teardown will usually say which. Establishing that the sag occurred and establishing that the sag caused the damage are entirely different evidentiary burdens, and only the second one decides the case.

Power quality evidence and its limits

The raw monitor data, not the summary report — waveform captures and the instrument's class under IEC 61000-4-30 determine what the record can support.
Monitor location, connection, and calibration record. A recorder connected on the wrong side of a transformer measures a different event than the one being argued about.
Utility event records and fault history for the feeder, which frequently explain the source of a sag the customer only saw the effect of.
Equipment nameplate, specification, and immunity rating — without the specification, susceptibility cannot be evaluated.
The failed equipment itself, for teardown. Attribution usually depends on the failure mode, not the event record.
Facility one-line diagram and the load configuration at the time of the event, including which equipment was running.
Timing correlation between the event record and the failure, to the resolution the records actually support.

Open tools and cross-checks

Where the suspect event is a voltage sag caused by starting a large motor rather than by a utility fault, the motor starting calculator at PowerEngCalc estimates the depth and duration of the resulting sag from the machine and source data, which is often enough to confirm or eliminate the on-site theory before a monitor is ever installed. The voltage drop calculator covers the steady-state case where the complaint is chronic low voltage rather than a discrete event.

Power Quality Investigation