What Power Factor Does to a Current Reading

Why this matters

Amps are the number every tech has. They are also the number that carries two things at once: the current doing work and the current doing nothing but going back and forth. A clamp cannot tell them apart, which makes a current reading completely honest about conductor heating and breaker loading and only partly honest about anything else.

The practical consequence is a pattern that gets misread on service calls every week: the amps went up and the work went down. A tech who has learned "over nameplate means working too hard" will replace a healthy machine on that reading, and the replacement will do exactly the same thing.

Before you clamp anything

On a roof, fall protection comes before diagnosis. The general industry duty for walking and working surfaces is at 29 CFR 1910.28, with a 4 ft trigger height, and construction work on the same roof falls under 29 CFR 1926.501 with a 6 ft trigger. A field-service shop can land under either depending on the job, so know which one your work is being performed under before you step off the ladder.

Clamp outside the enclosure wherever the conductors are reachable there, because that is a non-contact measurement and it costs you nothing.

A voltage reading requires contact. 29 CFR 1910.333(a)(1) requires live parts to be de-energized before work on or near them unless de-energizing introduces additional or increased hazards or is infeasible due to equipment design or operational limitations; the electrical lockout and tagging procedure is 29 CFR 1910.333(b)(2), since 1910.147 excludes exposure to electrical hazards from work on conductors and equipment in electric utilization installations at (a)(1)(ii)(C). Where the reading must be taken energized, prove the instrument live-dead-live under NFPA 70E-2021, 120.5 in the edition your employer's electrical safety program adopts, use the approach boundary and arc-flash PPE from that edition, inspect the leads for damaged insulation first, and confirm the meter's measurement category under the IEC 61010 series covers the point you are landing on.

The call, and the number that started it

A small commercial customer reports that a three-phase rooftop unit trips its overload on hot afternoons and runs fine in the morning. The compressor motor's nameplate full-load current is 22.4 A on a nominal 208 V supply.

The first tech clamps one leg in the afternoon, reads 24.6 A, writes "amps over nameplate, motor working too hard", and quotes a motor.

That reading is real. Every conclusion built on it is wrong, and the fastest way to see it is to check whether the number could have caused the symptom.

The arithmetic that kills the theory

An overload relay opens on current. Sizing for the separate overload device comes from NFPA 70, the National Electrical Code, Article 430, in the edition your authority having jurisdiction has adopted, which sets it at 115 or 125 percent of nameplate full-load current depending on the motor's service factor and temperature rise. On a 22.4 A nameplate that is 25.8 A at the lower setting and 28.0 A at the higher one.

The tech's 24.6 A sits below both. The current he measured could not have opened the relay at either permitted setting. Whatever tripped the unit, he had not found it, and the motor quote was written on a number that failed its own test.

That check takes fifteen seconds and it belongs in front of every "amps are high" conclusion. A current reading that does not reach the trip point of the device that tripped is evidence of something, but not of the trip.

Correcting the comparator

The second problem is the comparison itself. Nameplate full-load current is stated at rated voltage, rated frequency and rated shaft load. Read amps at any other voltage and you are comparing two numbers that do not share a basis.

For a motor holding a constant mechanical load, current rises roughly in inverse proportion to voltage over a modest deviation: the shaft still needs the same work, so as voltage falls, current climbs to supply it. That relationship is derived for a loaded induction motor over a small deviation from rated voltage, and it understates the effect slightly at larger sags because losses grow too.

The tech had a morning reading in the file: 23.2 A on the same leg, with 207 V line to line measured at the unit. His afternoon voltage measured 197 V, which is 5.3 percent below the 208 V nominal.

Expected afternoon current from the voltage change alone: 23.2 times 207 over 197, or about 24.4 A. Measured: 24.6 A. The residual is 0.2 A, which is under 1 percent of the reading and comfortably inside a field clamp's stated accuracy, so it cannot be resolved as anything.

The shaft load did not change. The voltage did, and the current followed it. The correct comparator was never 22.4 A. It was the current this motor should draw at 197 V for the load it was carrying, and against that comparator the machine was behaving normally.

The reading that found the trip

Clamping all three legs instead of one: 22.9 A, 24.6 A and 28.1 A. Average 25.2 A, and the worst leg deviates from that average by 2.9 A, which is 11.5 percent current unbalance.

The 28.1 A leg is at the 125 percent overload setting and well past the 115 percent one. That is the trip, and it was on the leg nobody clamped.

Measuring all three line-to-line voltages at the same moment: 197 V, 201 V and 194 V. Average 197.3 V, worst deviation 3.7 V, which is 1.9 percent voltage unbalance. The rule of thumb for a polyphase induction motor near full load is that percent current unbalance runs roughly six to ten times percent voltage unbalance, and 11.5 over 1.9 is about six, which puts this squarely inside the expected relationship rather than pointing at a motor fault.

So the fault is upstream: a supply unbalance, from the utility or from single-phase load distributed unevenly inside the building, appearing on hot afternoons when that single-phase load is largest. Replace the motor and the new one gets the same unbalanced supply, heats the same leg, and opens the same relay.

What power factor did while all this happened

Here is the part that catches people. At reduced voltage on a loaded induction motor, the magnetizing current falls roughly with voltage while the working current rises. The ratio of real to apparent power therefore improves. Power factor got better on the afternoon this unit was tripping.

Which is the whole lesson in one line. Power factor is a property of the load's character and its loading, not a health reading, and it moved in the flattering direction while the situation deteriorated. Current moved in the alarming direction while the shaft work stayed flat. Neither number, alone, described the machine's condition.

Direction table: when current lies about work

The last three are the ones that catch techs from the other direction, because they invert the intuition that a struggling machine draws more. State the geometry with them: the fan behaviour holds for centrifugal fans, and an axial fan pushed near stall does not follow the same curve; the pump behaviour holds for radial-flow centrifugal machines, and an axial-flow pump draws its highest power at shutoff, not its lowest.

What to record so an amp reading stays comparable

• All phases, not one. A single leg on a three-phase machine is a sample of a set that may not be uniform, and the unbalance is often the finding.
• Voltage at the same terminals at the same moment. Without it, the current has no comparator.
• Whether the meter is true RMS. On distorted waveforms an averaging-responding instrument and a true-RMS instrument legitimately differ, and the difference is not a fault.
• The operating point. Ambient, stage, damper or valve position, whatever sets the load. A compressor's current climbs with condensing temperature, so an amp reading with no ambient beside it cannot be compared to next month's.
• The overload setting actually installed, so the next person can run the fifteen-second check above instead of repeating the first tech's mistake.

How to verify you got this right

• Ask whether your number could have caused the symptom. If the device that opened responds to current, compare your measured current to its actual trip point before you build a theory on it.
• Clamp every leg before concluding anything on a polyphase machine.
• Take voltage and current in the same minute, because comparing an afternoon current to a morning voltage produces a correction that looks rigorous and is not.
• Repeat at a second operating point. If the current tracks voltage inversely across both points with the load unchanged, you have confirmed a supply effect rather than assumed one.
• Confirm the direction against the machine type before you use the table above. Fan and pump geometry decide the sign, and using the wrong sign turns a good diagnostic into a confident wrong answer.

References

• 29 CFR 1910.28 (general industry walking-working surfaces) and 29 CFR 1926.501 (construction fall protection)
• 29 CFR 1910.333(a)(1) and (b)(2), with the carve-out at 1910.147(a)(1)(ii)(C)
• NFPA 70E-2021, 120.5, as adopted through your employer's electrical safety program
• NFPA 70, National Electrical Code, Article 430, in the edition adopted by your authority having jurisdiction
• Motor manufacturer data for full-load current, service factor and permissible voltage unbalance; NEMA motor standards for unbalance derating
• See related: Measuring Electrical Power Versus Measuring Current; What Power Factor Actually Costs a Shop; Measuring Current Draw Against the Rating