Malaysian Power System Harmonic Analysis - Carelabs Malaysia
A harmonic analysis establishes how far the current and voltage waveforms in an installation depart from a sine wave, whether that distortion breaches the limits applicable at the point of common coupling, and which loads are responsible. It is a measurement and modelling exercise, and the sequence it follows determines whether the conclusions hold.
Distortion comes from loads that draw current in pulses rather than smoothly. Variable speed drives, UPS systems, rectifiers, LED lighting drivers, induction furnaces and switched-mode power supplies all do this. Each injects current at multiples of the 50 Hz fundamental, and that current flowing through the source impedance of the network produces voltage distortion that every other load on the same busbar then experiences.
The limits the work is assessed against
IEEE 519 is the framework used in most Malaysian industrial work. It sets two distinct classes of limit, and conflating them is the most common error in harmonic reporting.
Voltage distortion is the responsibility of the network. IEEE 519-2014 sets a total harmonic distortion limit of 8.0% for systems at or below 1 kV, and 5.0% for systems above 1 kV up to 69 kV, with individual harmonic limits below those.
Current distortion is the responsibility of the connected installation and is expressed as total demand distortion, referenced to maximum demand load current rather than to the instantaneous current. The permitted TDD depends on the ratio of short circuit current to maximum demand current at the point of common coupling. A stiff supply relative to the load is allowed more distortion; a weak one, less. For the lowest ratio band the limit is 5.0%.
Because TDD is referenced to demand, a lightly loaded plant can show alarming THD percentages while remaining compliant, and a heavily loaded one can show a modest THD figure while breaching TDD. Reporting THD alone tells you very little about your position.
How the study runs
Define the point of common coupling. Limits apply at the boundary with the utility, not at an individual drive panel. Fixing the PCC first is what makes the later compliance statement meaningful.
Measure under representative conditions. Distortion varies with load, so a snapshot at one moment answers the wrong question. Measurement is logged across a period that captures the real operating profile — typically a full week, covering production shifts, start-up, and the low-load periods when capacitor banks behave differently. Instrumentation meeting IEC 61000-4-30 and IEC 61000-4-7 records the harmonic spectrum rather than a single distortion figure.
Model the network. Measurement tells you the present position. A model tells you what happens when the network changes, and specifically where resonance sits. Power factor correction capacitors form a resonant circuit with the supply transformer inductance, and if that resonant frequency lands near a harmonic the plant actually produces — the 5th at 250 Hz and the 7th at 350 Hz being the usual candidates — the result is amplification rather than correction. Capacitors then fail, fuses operate for no apparent reason, and the cause is often misread as a capacitor quality problem.
Attribute the distortion. Spectrum analysis at individual feeders identifies which loads inject what. Six-pulse drives characteristically produce 5th, 7th, 11th and 13th harmonics; the pattern is recognisable and lets attribution be made on evidence rather than on assumption.
Assess and recommend. Results are compared against the applicable limits, and where mitigation is needed the options are set out with their trade-offs: line reactors and DC link chokes at the drive, passive tuned filters at the busbar, active filters where the load profile varies too much for a fixed tuning, or a phase-shifting transformer arrangement where harmonic cancellation is achievable by design.
Why the modelling step is not optional
A measurement-only report describes the plant as it is today. Most harmonic problems in Malaysian industrial facilities appear after a change — a new drive line, an uprated transformer, a capacitor bank retrofit — and the value of a calibrated model is that it answers whether a planned change will cause a problem before the equipment is bought. Running the study without it means every future modification is a fresh experiment on live plant.
A short circuit study underpins the same model, since the source impedance that governs resonance is the same quantity that governs fault current.
Frequently Asked Questions
Long enough to capture the plant's real operating profile, which in practice means a full week. Distortion varies with load, so a short snapshot risks recording either a production peak or an idle period and calling it representative. Where a plant runs distinct seasonal or campaign modes, measurement is repeated in each rather than extrapolated from one.
THD expresses distortion relative to the fundamental present at that moment, so it rises as load falls even when nothing has changed. TDD references distortion to maximum demand load current, which is why IEEE 519 uses it for current limits. A lightly loaded plant can show high THD and still comply; reporting THD alone does not establish a compliance position.
No, and they frequently make it worse. Capacitors form a resonant circuit with the supply transformer inductance, and if the resonant frequency falls near a harmonic the plant produces, the distortion is amplified rather than reduced. Detuned reactors or a properly tuned filter are the correct approach, and which one applies is a modelling question.
Need Expert Electrical Safety Support?
Carelabs provides arc flash studies, power system analysis, and MS IEC 60364 compliance services across Malaysia.