Neutral Overheating on a Retrofit Harmonic-Rich 4-Wire Riser
An IT park in Ahmedabad retrofitted its 40 W fluorescent fittings with LED panels and replaced old desktop UPS units with switched-mode power supplies. The 415 V, 3-phase, 4-wire rising main to the second-floor distribution board now reads, on a true-RMS clamp meter, line currents of 62 A, 60 A and 61 A with a current THD of 78 % on each phase. The neutral bar of the same cable reads 41 A. The cable is 4-core 70 mm² copper XLPE run in free air in a shaft at 44 °C, with the neutral sized at 50 % of the phase conductor, i.e. 35 mm². The landlord's main static meter has begun disagreeing with the tenant sub-meters by about 4 %. The maintenance team's proposal is to 'fit a bigger neutral and re-tighten the terminals'. No capacitor bank or filter exists on the installation.
1. Using the phase current and THD, estimate the fundamental and distortion current per phase, and explain why the neutral carries 41 A when the phases are nearly balanced.
Read the worked answer
Step 1 - split the RMS current. With THD_I = 78 % and I_rms = 61 A:
- I₁ = 61 / √(1 + 0.78²) = 61 / 1.268 ≈ 48.1 A fundamental
- I_distortion = √(61² − 48.1²) = √(3721 − 2314) ≈ 37.5 A
Step 2 - why the neutral is loaded. In a balanced 4-wire system the fundamental currents cancel vectorially (120° apart), so the neutral fundamental is ≈ 0. Triplen harmonics (3rd, 9th, 15th) are in phase in all three phases and add arithmetically: I_N = 3 × I₃.
Measured I_N = 41 A ⇒ I₃ ≈ 41/3 ≈ 13.7 A per phase.
Step 3 - interpret. If the whole 37.5 A of distortion were third harmonic, the neutral would read ≈ 112 A, i.e. nearly twice the phase current. The actual 41 A shows triplen content is only about a third of the distortion, with 5th and 7th dominating the rest. That is the key judgement: the neutral is not overloaded today, but the design has no margin if LED/SMPS share rises and the triplen component grows.
2. Is the 35 mm² neutral acceptable now and for the planned expansion? Support the answer with the derated rating and the applicable wiring-code rule.
Read the worked answer
Current capability of 35 mm² Cu XLPE, 3-core in free air: tabulated ≈ 110 A. Ambient correction at 44 °C for XLPE (Ca ≈ 0.87): 110 × 0.87 ≈ 96 A.
Against 41 A measured, the neutral is thermally fine today - roughly 43 % loaded.
But the rule, not just the arithmetic, governs the decision. IEC 60364-5-52 / IS 732 require that where the third-harmonic content of the line current exceeds about 33 %, the neutral be treated as a current-carrying conductor and sized equal to the phase conductor. Here THD is 78 %, so this installation already falls squarely inside that rule. Sizing the neutral at 50 % was a design error.
Worst case check: if the tenant load doubles and the neutral current reaches 3 × I₃ with I₃ ≈ 27 A, the neutral would carry ≈ 82 A - still under 96 A, but with no margin for the 5th/7th addition and with the terminal temperature already elevated by skin and proximity effects in the shaft. Recommendation: neutral = phase size (70 mm²) at the next cable change, and balance the single-phase SMPS load across phases as an interim measure.
3. The main meter and sub-meters disagree by 4 %. List, in priority order, what you would check and why harmonic distortion matters here.
Read the worked answer
Priority checks:
- CT ratio and meter programming - a wrongly entered CT ratio or a reversed CT polarity produces exactly this kind of small, persistent percentage error; verify primary/secondary ratio and the meter's programmed constant against the test certificate.
- Meter class and true-RMS capability - the main meter is a class 1 static meter that bills all harmonic power; many tenant sub-meters are class 2 or measure only the fundamental (or use a rectified-average front end). A 4 % gap is precisely the range produced by distortion power on an SMPS-dominated load.
- CT burden - check the VA burden of the pilot cable plus meter against the CT's rated burden; an over-burdened CT introduces ratio and phase errors at low load.
- Wiring/connection errors - cross-phased voltage and current inputs give a low but non-zero reading rather than a gross error.
Judgement: do not adjust the sub-meter readings to match. Prove the main meter first with a portable reference standard, then decide whether the discrepancy is a metering fault or a legitimate distortion-power billing effect.