Power Quality Events in Modern Electrical Networks

Abstract

Executive summary

Power-electronic loads and inverter-coupled generation have changed the disturbance environment faster than the monitoring infrastructure built to watch it. RMS-based metering is now structurally blind to the events that drive asset failure.

The proliferation of variable-frequency drives (VFDs), grid-tied PV inverters, EV chargers and UPS systems has altered the nature of disturbances imposed on shared electrical infrastructure, while grid-connected solar PV introduces both new harmonic sources and new network-dependent loss pathways. Conventional meters and SCADA polling — working at one-cycle RMS or slower — cannot resolve the sub-cycle switching transients, short-duration sags and swells, interharmonics and resonance phenomena that now dominate the failure modes of transformers, voltage transformers, motors and inverters.

This paper analyses each major disturbance category, quantifies its physical consequences on transformer insulation ageing, PV yield and plant assets, and sets out the measurement requirements a competent system must meet under IEC 61000-4-30, IEC 61000-4-7, EN 50160 and IEEE 519-2022. Two field case studies — an Indian industrial facility and a 110 kV transmission point monitored by the Janitza UMG 800 — demonstrate the principles in practice.

Standards framework

Applicable standards and their roles

Power quality is governed not by a single document but by a layered framework — measurement methods, emission and compatibility limits, equipment immunity, and national adoption. The table answers, at a glance, which standard governs what. Each is applied in its current published edition.

Regulatory landscape — India

Indian regulatory perspective

In India, power quality is governed by a three-layer framework — statutory regulation, national standards, and state-level rules. It is maturing rapidly but is not yet uniformly enforceable, which shapes how asset owners should approach monitoring.

Statutory layer — CEA and CERC

The Central Electricity Authority (CEA) sets the binding technical baseline through the CEA (Technical Standards for Connectivity to the Grid) Regulations, 2007 and the CEA (Grid Standards) Regulations, 2010, while the CERC Indian Electricity Grid Code (IEGC) fixes permissible voltage variation at grid nodes — tightening with voltage level (broadly ±5% at 765/400 kV, widening toward +11%/−10% at 220 kV). Electrical safety and supply obligations fall under the CEA (Measures relating to Safety and Electric Supply) Regulations.

National standard — IS 17036

India's first dedicated power-quality standard, IS 17036:2018 — Distribution System Supply Voltage Quality, was notified by the Bureau of Indian Standards in October 2018; the IEC 61000 EMC series is adopted nationally as the IS 14700 series (for example, IS 14700-4-30 ≡ IEC 61000-4-30). BIS continues to develop power-quality indices, assessment methods and measuring-equipment standards through its electrotechnical committee.

State rules and engineering practice

State Electricity Regulatory Commissions are progressively issuing dedicated PQ regulations — for example the Kerala SERC (Power Quality for Distribution System) Regulations, 2019, applicable at and below 33 kV — and the Forum of Regulators has published model PQ regulations to encourage wider adoption. In Tamil Nadu, demand-side management and supply-quality obligations sit with TNERC and the distribution utility. In practice, enforceable harmonic and event limits with continuous-monitoring obligations remain uneven across states. Until they are uniform, the asset-protection argument of this paper holds independently of compliance status: industrial and utility asset owners benefit from treating continuous, high-resolution monitoring at equipment terminals as an engineering necessity rather than waiting for a statutory mandate.

Section 01

The changing nature of power-network disturbances

IEC 61000-4-30:2015+AMD1:2021 Cl. 3 defines power quality as the characteristics of the electricity at a point on the system, evaluated against reference technical parameters. Three concurrent developments have transformed that environment since the early 2000s:

• IGBT power electronics everywhere. VFDs, grid-tied inverters and switched-mode supplies switch at 2–20 kHz, generating high-frequency spectra entirely outside the bandwidth of conventional measurement.
• Distributed renewable generation. Solar PV adds new harmonic sources and new sensitivity — inverter MPPT algorithms react directly to voltage disturbances at the point of common coupling (PCC).
• Growth in capacitor compensation. Shunt capacitance interacting with transformer leakage inductance and cable capacitance forms high-Q resonant circuits that amplify specific harmonics to damaging levels.

The result is a network that is simultaneously more disturbance-rich and more sensitive than at any previous point in its history. Monitoring designed for an earlier era is structurally inadequate for it.

Section 02

Switching transients — the invisible sub-cycle stressor

2.1 Definition and sources

A switching transient is a short, oscillatory or impulsive voltage excursion superimposed on the fundamental. IEC 61000-4-30 Cl. 3.21 classifies transient overvoltage events; IEC TR 60071-4 gives insulation-coordination context. Industrial sources include capacitor energisation (oscillatory inrush 2–50 kHz), VFD IGBT commutation notches, transformer magnetising inrush, PV inverter PWM switching, and arc-furnace electrode movement.

2.2 The measurement challenge

A 200 µs transient changes the computed 10-cycle RMS value by less than 1% — statistically invisible to any instrument relying on cycle-by-cycle RMS integration.

2.3 Consequences — insulation overstress and partial discharge

Voltage spikes exceeding the basic insulation level (BIL) intensify the local electric field within winding structures. Per IEC 60076-3:2013, even sub-BIL stress, repeated thousands of times weekly, initiates partial-discharge (PD) activity that erodes paper insulation through chemical decomposition. The failure pathway — transient → PD ignition → progressive erosion → delayed inter-turn failure — may span months with no conventional alarm; dissolved-gas analysis per IEC 60567 is the only post-hoc indicator. Fast fronts also distribute non-uniformly: the first 10–15% of turns absorbs 2–3× its rated share of the initial voltage gradient (distributed LC-ladder behaviour), concentrating turn-to-turn breakdown risk.

Section 03

Voltage sags — mechanical impulse loading and process disruption

3.1 Definition and classification

Per IEC 61000-4-30 Cl. 3.3, a voltage dip (sag) is a temporary reduction in RMS voltage below a threshold — typically 0.9 pu of declared voltage (U_declared) — characterised by retained voltage and duration. EN 50160 Annex A classifies events on the UNIPEDE DISDIP matrix; IEC 61000-4-11:2020 defines equipment-immunity test levels at 0%, 40% and 70% residual voltage.

3.2 Consequences on power transformers

Each sag followed by rapid recovery imposes a mechanical impulse on winding conductors proportional to current × flux density. Per IEC 60076-5:2006, winding integrity is designed for specific fault-current magnitudes — not the cumulative effect of repeated recovery inrush. Over hundreds of sag-recovery cycles, inter-conductor spacers and conductor geometry degrade, eventually enabling inter-turn fault under normal stress — a failure mode with no thermal precursor.

3.3 Consequences on solar PV plants

Grid-connected PV inverters follow fault-ride-through requirements set by IEC 61727:2004 Cl. 5.2 together with modern grid-interconnection standards (IEEE 1547-2018, IEC 62116, IEC 62109). During sags approaching the LVRT threshold, MPPT resets each cost 0.3–2 s of sub-optimal extraction; anti-islanding protection per IEC 62116:2014 may false-trip on deep sags, each followed by a 20–60 s reconnection delay.

3.4 ITIC / CBEMA classification

The ITIC 2000 voltage-tolerance envelope is widely used alongside IEC 61000-4-11 immunity levels. In this paper, events below the ITIC lower boundary at any duration are flagged as a region where equipment mal-operation or damage is expected — actionable engineering assessment beyond a pass/fail compliance metric.

Section 04

Voltage swells — dielectric overstress and accelerated ageing

4.1 Definition

IEC 61000-4-30 Cl. 3.4 defines a swell as a temporary RMS increase above a threshold, typically 1.1 pu. EN 50160 Cl. 4 requires the 10-minute mean RMS to stay within ±10% of nominal for 95% of any one-week period. Causes include large inductive-load disconnection, asymmetric fault clearing, ferro-resonance, and PFC over-compensation.

4.2 Transformer thermal ageing — IEC 60076-7 Arrhenius framework

IEC 60076-7:2018 codifies the Arrhenius ageing model for paper-oil insulation. Relative ageing rate V is exponential in hot-spot temperature θ_H, with the simplified rule that V doubles for each 6 °C rise above the reference (98 °C conventional paper / 110 °C thermally upgraded). Sustained swell raises core flux density, increasing core and winding losses and shifting thermal equilibrium upward.

4.3 Consequences on solar PV inverters

Grid-connected inverters disconnect when supply exceeds 1.1 pu for more than 0.5 s per IEC 61727 Cl. 5.3 and the applicable grid code. Repeated swell exposure additionally stresses DC-link capacitors, IGBT gate-oxide layers and EMC filter capacitors. Manufacturer MTBF assumes supply within the EN 50160 normal range; a high-swell environment materially reduces actual service life.

Section 05

Harmonic distortion — thermal stress and network resonance

5.1 Measurement standard and assessment framework

IEC 61000-4-7:2002+AMD1:2008 defines the method: harmonics from a 10-cycle (200 ms) windowed DFT, grouped into 10-minute statistics per IEC 61000-4-30. THD-V = √(Σ U_n²)/U₁ × 100% for n = 2…50. Harmonic voltage limits are voltage-level dependent: EN 50160 Cl. 4.4 covers LV and MV public networks (≤ 36 kV in India), while IEEE 519-2022 Table 1 defines PCC limits across all levels and must be applied in preference for HV (> 36 kV). Underlying these limits are the compatibility levels of IEC 61000-2-2 (public LV systems) and IEC 61000-2-4 (industrial-plant environments) — the reference disturbance levels against which emission and immunity are coordinated, and the basis for classifying an industrial electromagnetic environment (Class 1–3).

5.2 Core-loss increase — frequency dependence

Eddy-current loss components increase approximately with the square of frequency; total transformer losses, however, also include stray, structural and winding components and are more complex than a single f² term. On the eddy component alone, the 5th harmonic (250 Hz) contributes roughly 25× per unit voltage and the 7th (350 Hz) roughly 49× relative to the fundamental. Where 5th-harmonic voltage reaches 3–4% (within EN 50160 limits), its contribution to core eddy loss can be 10–15% of the fundamental-frequency eddy loss — invisible to THD-V monitoring without spectrum resolution.

5.3 Winding eddy loss and K-factor

Harmonic currents raise winding eddy loss via the K-factor: K = Σ (I_h/I₁)² × h² for h = 1…h_max. K = 1 is sinusoidal; K = 4–13 is typical for VFD-dominated industrial loads. Operating a K = 1 rated transformer at K = 4 without derating raises hot-spot temperature measurably, accelerating ageing per IEC 60076-7. K-factor derating per IEEE C57.110:2018 requires measurement of all significant harmonic currents contributing to eddy-current losses.

5.4 Network resonance and amplification

Parallel resonance between transformer leakage inductance (L_T) and system shunt capacitance (C_sys) occurs at f_r = 1 / (2π √(L_T·C_sys)). When f_r coincides with an emitted harmonic, that harmonic voltage is amplified by the circuit quality factor Q (5–30 in practice). A 5th-harmonic current injection of 5% may produce 25–50% 5th-harmonic voltage at resonance — causing sustained overheating even when aggregate THD looks acceptable (IEC TR 61000-3-6:2008 Cl. 4.3).

Section 06

Interharmonics and supraharmonics — the emerging frontier

6.1 Interharmonics

Interharmonics are components at non-integer multiples of the fundamental. IEC 61000-4-7 Cl. 6 defines subgroup measurement. Sources include cycloconverters, arc furnaces and non-synchronous PWM topologies. They cause light flicker (via subharmonic modulation), interference with power-line communication, and subsynchronous resonance — none captured by THD-V or standard harmonic analysis.

6.2 Supraharmonics (2–150 kHz)

Supraharmonic emissions from IGBT inverters — PV, EV chargers, VFDs — in the 2–150 kHz range are addressed by IEC TR 61000-2-2 and the developing IEC 61000-4-19 measurement standard.

Section 07

Voltage interruptions, flicker and frequency deviations

7.1 Voltage interruptions

IEC 61000-4-30 Cl. 3.5 defines an interruption as voltage below 0.01 pu (1%) of declared. Short interruptions (< 3 min) are distinguished from long. The recovery transient following automatic reclosure may be the single most severe transient stress in a transformer's supply history.

7.2 Voltage flicker (P_st, P_lt)

Fluctuation at 0.5–25 Hz causes perceptible flicker, characterised by short- and long-term perceptibility indices per IEC 61000-4-15:2010. EN 50160 specifies P_lt ≤ 1.0 for 95% of a one-week period. For transformers, flicker-causing loads (arc furnaces, rolling mills) mechanically cycle winding conductors through rapid current swings.

7.3 Frequency deviations

IEC 61000-4-30 Cl. 3.2 requires frequency resolution of at least 10 mHz. EN 50160 specifies ±0.5 Hz of 50 Hz for 95% of any one-week period. For PV inverters, grid frequency is the PLL reference; excursions can destabilise the PLL and trigger nuisance anti-islanding trips per IEC 62116:2014.

Section 08

Why IEEE 519 / EN 50160 compliance alone is not sufficient

A common misconception equates regulatory compliance — IEEE 519 PCC voltage limits and EN 50160 weekly statistics — with adequate asset protection. Compliance is necessary but structurally insufficient.

8.1 The PCC measurement limitation

IEEE 519-2022 Cl. 5 sets harmonic-voltage limits at the PCC — the utility-customer boundary. A PCC measurement characterises the installation's aggregate contribution to the shared network, not the internal environment at equipment terminals. Transformer secondary busbars, MCC busbars and VFD output terminals may see harmonic voltages 2–5× higher than the PCC due to internal resonance. A compliant PCC reading says nothing about internal conditions.

8.2 The 95th-percentile illusion

Both standards evaluate the 95th percentile of 10-minute values over one week. That means 5% of intervals — 504 minutes (8.4 hours) per week — may legally exceed limits with no non-compliance finding. Where resonance or sag events cluster in specific production shifts or switching sequences, the 5% window coincides precisely with the highest-consequence periods. Statistical compliance offers no protection against systematic worst cases. This is not a deficiency in IEEE 519 or EN 50160 — percentile evaluation is deliberate and well-suited to network planning — but an inherent property of statistical assessment that asset owners must understand.

8.3 The duration problem

Sub-second events — transients (0.3–5 ms), sags (20–200 ms), resonance bursts (50–500 ms) — are by definition absent from 10-minute statistics, yet they are the primary drivers of insulation stress, MPPT disruption and nuisance trips. A 10-minute average THD-V of 4.5% (within EN 50160) is fully consistent with 20 capacitor-switching transients per hour at 1.8 pu peak, accruing dielectric damage with no statistical footprint.

8.4 The continuous-monitoring imperative

Asset protection requires continuous monitoring with sufficient temporal resolution to catch every disturbance category — not periodic surveys. An adequate system must simultaneously: (a) measure in accordance with IEC 61000-4-30 for all covered parameters at high accuracy; (b) sample waveforms fast enough to detect sub-cycle events at the network's disturbance spectrum; and (c) be permanently deployed at equipment terminals — not only at the PCC.

Section 09

Power-transformer ageing — a multi-mechanism failure model

PQ-driven transformer failures rarely have a single proximate cause. The reality is a multi-mechanism process — thermal ageing, PD initiation, mechanical fatigue and dielectric failure — whose pathways interact synergistically.

9.1 Insulation thermal ageing — IEC 60076-7

9.2 Partial discharge — the silent pathway

PD is localised breakdown that does not bridge the full gap. It erodes paper through chemical decomposition, generating dissolved gases detectable by DGA per IEC 60567. The key insight is initiation: switching transients momentarily intensify the field at conductor edges and oil-paper interfaces, igniting PD that then propagates under normal voltage — weeks or months after the initiating events. This temporal decoupling makes causal attribution impossible without historical transient waveform data.

9.3 The multi-mechanism monitoring gap

SCADA at 15-minute polling and analysers computing 10-cycle RMS cannot detect the sub-cycle phenomena that drive degradation: a 500 µs transient leaves no RMS trace; a 40 ms sag may fall entirely between SCADA intervals. Conventional SCADA frequently fails to detect the initiating PQ phenomena responsible for long-term degradation — so by the time installed infrastructure flags a problem, insulation stress is often already advanced.

Section 10

Solar PV plant efficiency loss

Grid-connected PV is exposed to the full disturbance spectrum from both the grid and the plant's own power electronics, with distinct loss pathways for each category.

10.1 Inverter MPPT disruption

Voltage sags, swells, transients and rapid variations perturb the DC bus through the control loop, forcing MPPT resets. Re-acquisition typically takes 0.3–2 s. For a 100 kWp string inverter, 10 sag events/day at 1 s re-acquisition is ≈1 hour of degraded yield annually — invisible to energy-meter recording.

10.2 Harmonic-induced yield reduction

High THD-V at the AC terminals forces the current controller to inject compensating harmonics, raising switching losses and lowering conversion efficiency.

Section 11

Consequences on other critical assets

11.1 Voltage transformers — dielectric failure

Electromagnetic VTs are especially vulnerable: their primary windings present high impedance to high-frequency transients, concentrating voltage across insulation near the primary terminal. First-turn insulation may see 2–3× rated gradient during fast fronts; repeated sub-threshold stress initiates PD and leads to failure weeks after the causative pattern.

11.2 VFDs — DC-bus capacitor degradation

Harmonic current flows preferentially through low-impedance DC-bus capacitors, raising I²R losses and temperature and accelerating electrolytic degradation. A sustained input THD-I of 30–40% (common without input reactors) is reported to reduce DC-bus capacitor life by up to ~50% relative to sinusoidal supply lit.

11.3 PFC capacitor banks

Fixed PFC banks present low impedance to harmonics, concentrating harmonic current in the elements. IEC 60831-1:2014 sets maximum permissible total RMS current at 1.3× rated. Near the bank's resonant frequency, overload and premature failure are common — and invisible to conventional power-factor monitoring.

Section 12

Monitoring — technical requirements for a compliant system

The requirements below are not arbitrary; they follow from disturbance physics and the standards. Any system for asset protection must satisfy all of them. To be clear, conventional RMS metering remains entirely adequate for energy management, billing and basic operational supervision — the requirements here concern the distinct task of transient, event and asset-protection analysis, for which RMS-only measurement is insufficient.

12.1 A compliant system must

• Measure in accordance with IEC 61000-4-30 for all covered parameters (RMS, sags, swells, interruptions, harmonics, unbalance, frequency) at high accuracy and resolution. The standard defines Class A and Class S performance classes; the higher the accuracy, the stronger the contractual and asset-protection basis.
• Measure the harmonic spectrum continuously to at least the 50th harmonic (2.5 kHz) per IEC 61000-4-7 grouping, with 10-minute 95th/99th-percentile statistics per IEEE 519-2022.
• Acquire waveforms at ≥ 1024 samples/cycle for indicative transient characterisation, with pre/post-trigger capture.
• Provide per-cycle (10-cycle) RMS for sag/swell detection per IEC 61000-4-30 §5.4.2 / §5.5.
• Measure voltage unbalance (negative/positive-sequence) per §5.8 with 10-minute aggregation.
• Compute supply-availability metrics (MTBF/MTTR of PQ-threshold-exceedance events — distinct from equipment failure in the IEC 60050-191 sense).
• Support distributed deployment at equipment terminals (LV busbars, capacitor switching points, inverter outputs) — not only the PCC.

Section 13

High-resolution distributed monitoring in practice

One example of a monitoring platform meeting the requirements above is the Janitza UMG 800 series — a top-end, high-resolution power quality analyser in a compact DIN-rail form factor, engineered for cost-effective distributed deployment.

13.1 Measurement architecture

The UMG 800 acquires voltage and current at 1024 samples/cycle — ≈51.2 kHz at 50 Hz, ≈19.5 µs per sample — measuring to the IEC 61000-4-7 and IEC 61000-4-30 methodologies, with voltage measurement accuracy of 0.2% per the Janitza specification.

Source: Janitza UMG 800 technical data and datasheet (janitza.com). Specifications subject to change; verify against current Janitza documentation for contractual use.

Section 14 · Field evidence — Case study 1

Capacitor-bank resonance at an industrial facility — Field Case Study 1

Sections 1–13 set out the disturbance mechanisms and why high-resolution monitoring is required. The two studies that follow demonstrate those mechanisms in the field — first the issue as it presented, then the measured evidence and root cause. Both are actual field investigations recorded with the Janitza UMG 800; identifying details have been removed or generalised where necessary, and they are presented as representative engineering cases, not statistical samples.

14.1 The issue as presented

A Coimbatore manufacturing facility — 2 MVA, 11/0.415 kV transformer supplying 450 kW VFDs, 200 kW welding and 300 kW resistive heating, with 400 kVAr fixed PFC — suffered three VT failures, unexplained transformer overheating and VFD nuisance trips over 18 months. Installed SCADA showed THD-V 4.8% (within EN 50160), 72% load, PF 0.91. No corrective action was identifiable.

14.2 UMG 800 deployment and root cause

Three UMG 800 analysers were deployed at the 11 kV busbar, transformer LV busbar and PFC switching point. Within 72 hours:

• Repetitive 8–10 kHz transients, peak 1.8 pu, with every capacitor energisation (12–18/hour).
• Transient duration 300–800 µs — invisible to installed monitoring (per-cycle RMS error < 0.3%).
• Calculated parallel resonance from leakage inductance (4.2 mH) and system capacitance: 8.3 kHz — matching the observed oscillation precisely.
• Supply-availability MTBF (PQ-threshold events): 1 h 47 min — below the 99% target.

Mitigation (detuned reactors on the capacitor steps, shifting the bank's series-resonant point below the 5th harmonic) eliminated the resonance condition and the recurring VT failures.

Section 15 · Field evidence — Case study 2

Progressive single-phase fault on a 110 kV feeder — Field Case Study 2 (events #8380 / #8381)

Two event records were captured at a 110 kV point, each 20,000 samples at 51,229 Sa/s spanning 390 ms with an 80 ms (4-cycle) pre-trigger. All values in primary kV. Pre-event voltage was ≈3.7% above nominal at both events — within the EN 50160 ±10% range.

15.1 Event #8380 — 06 Apr 2026, 22:10:41 IST — single-phase reduction

Pre-event L2 reference 65.95 kV L-N (103.8% of nominal); dip threshold 59.35 kV. L2 declined over 4 cycles to a minimum of 61.42 kV (93.1% of reference) at cycle 8, then recovered to a new steady level of 63.68 kV — a permanent −3.4% step relative to L1/L3.

15.2 Event #8381 — 07 Apr 2026, 00:41:27 IST — significant sag

Pre-event L2 reference 65.384 kV L-N (102.9%); dip threshold 58.846 kV. Both events affect L2 exclusively within a 2.5-hour window — a clear pattern of a deteriorating or intermittent L2-feeder fault. L1 and L3 are undisturbed throughout.

15.3 Root-cause analysis

15.4 Recommendations

Immediate. Cross-correlate L2-feeder relay and SCADA fault logs at 06 Apr 22:10 and 07 Apr 00:41 IST; inspect L2 cable joints, line sections, tower earths and terminations for tracking/PD/corona; run a Tan-δ test on the L2 cable per IEC 60247 / IS 10810; investigate the permanent −3.4% L2 step (VT secondary wiring, primary connections, tap-changer logs).

Monitoring & mitigation. Set the UMG 800 trigger to 85% on L2 with a negative-sequence alarm at V2 > 2%; run a ≥7-day continuous assessment per IEC 61000-4-30; perform PD monitoring per IEC 60270; assess downstream motor impact per IEC 60034-26 (rotor heating ∝ (V2/V1)²). If L2 insulation failure is confirmed, replace the section per IEC 60502-2 / IS 7098; if sags recur, evaluate a Dynamic Voltage Restorer at the PCC.

Section 16

Limitations of the study

This paper is a standards-referenced engineering analysis intended for practical asset protection. Its scope and assumptions carry the following limitations:

• Case studies are illustrative. Both are real field investigations, but outcomes are site-specific and not statistical samples; results should not be generalised without site-specific measurement.
• Economic and yield impacts are site-dependent. PV loss figures, transformer life cost and similar values vary with load profile, climate, configuration and tariff. Figures marked lit. are indicative literature estimates, not standardised values.
• Life estimates require thermal modelling. Ageing figures (Table 2, §4.2) are first-order indicators. Contractual life assessment shall use the full IEC 60076-7 thermal model with site loading, ambient, cooling mode and insulation system.
• Supraharmonics need dedicated instruments. Content above the conventional harmonic range (≈2 kHz) measured at 1024 samples/cycle is diagnostic, not a compliant IEC 61000-4-19 assessment.
• Transient fidelity is bounded. 1024 samples/cycle gives detection and envelope capture; full fast-front impulse reconstruction (rise < 5 µs) requires IEC 61000-4-5 recorders.
• Monitoring complements, not replaces, engineering analysis. Continuous high-resolution data is necessary but not sufficient; root-cause attribution and mitigation still require qualified engineering judgement.

Section 17

Conclusion

The contemporary network's disturbance environment differs fundamentally from the one conventional monitoring was designed for. Switching transients, short-duration sags and swells, high-order and interharmonic components, and network resonance now dominate transformer ageing, equipment failure and PV loss — and they share one trait: they are sub-cycle, sub-second events invisible to any system based on fundamental-frequency RMS.

IEEE 519 and EN 50160 compliance is necessary but not sufficient. PCC-only measurement misses internal resonance; 95th-percentile statistics permit ≈8.4 hours of weekly exceedance; 10-minute aggregation omits the sub-second events that dominate degradation. Effective protection requires permanent, distributed, high-resolution measurement per IEC 61000-4-30 at equipment terminals — not only at network boundaries.

The 110 kV case study makes the point concrete: a progressive single-phase fault developing over 2.5 hours produced an ITIC-prohibited event at 64.8% retained voltage and 14.08% negative-sequence unbalance — 7× the EN 50160 limit — entirely missed by conventional infrastructure. The UMG 800's complete waveform record, symmetrical-component analysis and ITIC classification supplied the evidence required for definitive root-cause determination and targeted action.

Appendix

Normative references