Neutral conductor overheating in a three-phase four-wire system is a zero-sequence problem, so measure harmonic orders first and only then select a filter. A neutral that runs hot while the phases stay comfortably cool is rarely a coincidence. It points to zero-sequence current driven by triplen harmonics, which behaves nothing like a balanced-load calculation predicts.
This article follows the order an engineer should: the pain point, the cause, the thermal consequence, then measurement, and only then the choice between an active harmonic filter, an active load balancer and a static var generator.
The usual complaint is a neutral busbar, cable or transformer neutral terminal running warmer than any phase conductor. Terminations darken, the insulation near a lug hardens, and occasionally a thermal-magnetic breaker trips on a feeder that shows no apparent overload. Phase current readings look reasonable, so the neutral is assumed undersized and replaced with a larger conductor.
Upsizing sometimes lowers the temperature, which is why the mistake survives. The extra copper spreads the same losses over more cross-section, but the current producing the heating is untouched. If the source is a harmonic pattern that adds in the neutral instead of cancelling, a bigger neutral only defers the problem at extra cost.
Modern single-phase electronic loads — switch-mode power supplies, LED drivers, IT equipment, small drives and single-phase EV chargers — draw current in short pulses rather than smooth sine waves. Those pulses contain many individual harmonic orders, and in a three-phase four-wire system the behaviour of each order depends on its sequence.
The fundamental is positive-sequence. Harmonics of order 5 and 7 are negative-sequence, 11 and 13 are positive-sequence again, and in a balanced system their vector sums in the neutral are close to zero. Triplen harmonics — the third, 9th and 15th, every order that is an odd multiple of three — are zero-sequence. Their phase currents are in phase, so they do not cancel in the neutral. They add.
That sum is the zero-sequence current. It flows in the neutral and returns through the earth path and transformer, and can push neutral current above the phase current that supposedly defines the circuit rating. A neutral sized at the phase rating, assuming balanced phases cancel, is exposed to a current it was never dimensioned to carry.
Zero-sequence current produces losses through the same I²R relationship as any other current, but the concentration differs. Because the triplen currents from all three phases share one conductor, the neutral carries their arithmetic sum while each phase carries only its own share. The result is disproportionate neutral heating, plus additional eddy-current and stray losses in transformer windings, tanks and enclosures.
The effects compound over time. Insulation ages faster, terminations loosen as they cycle thermally, protective devices drift or trip, and transformer derating may become necessary. Voltage waveforms flatten, adding losses in motors and other equipment. None of this appears on a phase-current meter; it becomes visible only when neutral current and individual harmonic orders are captured. An active harmonic filter is one of the devices that can act directly on this harmonic current once it is quantified.
| Condition | Consequence | Measurement |
|---|---|---|
| Third harmonic | Zero-sequence current in neutral | Neutral current, 3rd order |
| 9th, 15th triplens | Extra winding and stray losses | Individual orders to 25th |
| 5th and 7th harmonics | Cancel in neutral | Per-phase THDi |
| Neutral near phase current | Insulation ageing, loose terminals | Neutral temperature, CT loading |
Measurement comes before selection. Two quantities drive the decision: the individual harmonic orders present, and the zero-sequence current the neutral actually carries.
A true-RMS power quality analyser with harmonic capability is the minimum instrument. It should record individual harmonic orders well beyond the 15th — to the 25th or higher where possible — with per-phase THDi and, critically, the neutral current. Many general-purpose meters report only total distortion and miss the triplen content that matters.
CT placement decides whether the data means anything. The neutral CT must sit on the neutral of the same three-phase four-wire circuit, not be inferred from phase readings, and the phase CTs must be at the intended point of common coupling. Ratio, class and burden should suit the expected neutral current, which may exceed phase current. If the neutral CT is omitted, undersized or allowed to saturate, the single most important number in the survey is absent.
Record over a representative period that includes the worst case — shift patterns, evening peaks, weekend profiles — and log power factor and reactive power with the harmonics. A spot reading on a quiet afternoon rarely captures the condition that overheats the neutral.
Once the data is in hand, selection follows from the dominant disturbance rather than from habit. If the neutral is hot and triplen harmonics dominate, the answer is a device that cancels those harmonic currents or rebalances them away from the neutral conductor.
An active harmonic filter injects compensating currents to cancel selected harmonic orders, lowering THDi and the triplen content that loads the neutral. An active load balancer works on the sequence components directly, redistributing current between phases to reduce negative-sequence and zero-sequence components, including the neutral current itself. A static var generator addresses reactive power and power factor, supporting voltage and trimming reactive current; it is not primarily a harmonic or neutral-current device.
In practice the three are often complementary. The active power filter category groups the options, while the device pages explain harmonic cancellation, neutral rebalancing and reactive compensation.
| System / load type | Likely solution | Selection data |
|---|---|---|
| Four-wire, switch-mode loads | Active harmonic filter (AHF) | Neutral current, per-order harmonics |
| Imbalanced single-phase distribution | Active load balancer (ALB) | Phase, neutral, sequence components |
| Low power factor, low distortion | Static var generator (SVG) | Reactive power, target power factor |
| Harmonics plus poor power factor | AHF + SVG | THDi, orders, reactive power |
Whichever route is chosen, the same inputs decide the rating: neutral current and per-order harmonic currents at the PCC, transformer rating and impedance, the load profile, the target power factor and the presence of capacitor banks. Without those, a proposal is a guess.
Active devices have limits that honest specification acknowledges. Compensation bandwidth is finite, so very high-order harmonics may remain; current rating caps how much disturbance a unit can absorb; and an undersized neutral or a loose termination is a hardware fault no filter corrects.
Resonance is the most common hidden risk. Where fixed or switched capacitor banks are installed, network impedance may resonate near a harmonic order and amplify distortion instead of reducing it. That interaction must be checked before adding reactive compensation, and detuning or a different strategy may be required.
Then there is maintenance and return. An active device needs ventilation, filter cleaning and periodic verification of its own performance, because a unit that has stopped compensating still looks healthy. Its economic case rests on the site's tariff, loss profile and load growth, not on a promised saving. The defensible approach is to quantify existing loss and distortion, set a measurable target, and verify after commissioning.
Yes. In a three-phase four-wire circuit feeding non-linear single-phase loads, the neutral can carry a current comparable to or larger than the phase current, because triplen harmonics add rather than cancel. Size the neutral for that zero-sequence current, and verify existing neutrals by measurement.
Phase CTs belong at the point of common coupling where the assessment is defined, and a CT is required on the neutral of that same circuit. Placing the neutral CT elsewhere, or relying on calculated rather than measured neutral current, removes the most relevant data point. Check ratio, class and burden against the expected neutral current, including its harmonic content.
Match the device to the dominant disturbance. Harmonic current and THDi point to an AHF; phase imbalance and neutral current point to an ALB; reactive power and power factor point to an SVG. Where several problems coexist, a combination is common, and the choice should follow measured data rather than rule of thumb.
Budget for ventilation, filter replacement and periodic performance verification, since an active device that silently stops compensating still looks healthy. The financial case depends on the site's measured losses, tariff and load growth; build it from measured data and verify it after commissioning, without fixed savings claims.
Start with a measurement, not a quotation. Capture neutral current and individual harmonic orders at the PCC, confirm whether triplen harmonics dominate, then match the data to a compensating strategy. To have a survey reviewed for a specific installation, send the measured neutral current, THDi and per-order harmonic figures to the YT Electric engineering team.
suscríbase a nosotros para disfrutar de los precios del evento y obtener algunos de los mejores precios.
compatible con la red ipv6