A system-boundary standard, enforced at one point only
IEEE Std 519-2022 was approved on 13 May 2022 and published in August 2022; it supersedes IEEE 519-2014. The document was renamed from "Recommended Practice and Requirements…" to "IEEE Standard for Harmonic Control in Electric Power Systems" — a deliberate signal that its limits are written to be applied as requirements, not advisory practice.
The standard sets steady-state harmonic limits at the interface between the supply system and the user — the Point of Common Coupling (PCC). It does not address transients, and it does not set limits for individual machines, panels, or feeders inside a facility. Compliance is a property of the boundary, established by a single statistical measurement campaign at that boundary.
Keep injected current harmonics (assessed as TDD and per-order limits) within Table 2/3/4, selected by the short-circuit ratio at the PCC.
Keep supply voltage distortion within Table 1 by managing system impedance — a near-sine voltage is owed to every connected user.
Indian regulatory context
Under the CEA (Technical Standards for Connectivity to the Grid) (Amendment) Regulations, 2019, the limits of voltage harmonics by the distribution licensee, the limits of current-harmonic injection by bulk consumers, and the point of measurement (the PCC) are to follow IEEE 519 "as amended from time to time." Because 519-2022 supersedes 519-2014, the 2022 edition is the operative reference, and TNERC implements the obligation in its power-quality framework. Measurement instruments must additionally satisfy IEC 61000-4-30 Class A (Ed. 3) with harmonic measurement per IEC 61000-4-7 — the BIS-recognised IEC measurement basis.
Definitions, exactly as 519-2022 frames them
Most disputes are not about harmonics at all — they are about getting one of these five quantities wrong. Two of them, IL and Isc, are external numbers inserted into the measurement, and they decide the verdict before the meter is even read.
The point on the public supply electrically nearest the load at which other loads are, or could be, connected — the interface between the system owner or operator and the user, located upstream of the installation. The PCC is a technical boundary, not a property of the DISCOM: it is the consumer's power-receiving point from the DISCOM, at which the consumer carries the responsibility to share its distortion data. It should not be equated with the DISCOM's metering point — doing so would affect the consumer's rights. For an HV consumer with its own transformer it lies on the HV / incoming side; for an LV consumer fed from a shared distribution transformer it is that transformer's LV side.
Isc is the actual available short-circuit current at the PCC, set by source MVA and the series impedance to the PCC. SCR = Isc/IL selects the current-limit band — a higher SCR (stronger grid) earns more relaxed limits.
RMS of the harmonic currents (orders up to the 50th, interharmonics excluded) as a percentage of the fundamental current. It rises at light load and falls at heavy load. It is not the compliance metric.
RMS of the harmonic currents (up to the 50th) as a percentage of the maximum demand current IL. This is the current-compliance metric. iTHD is only a step on the way to it.
Established at the PCC as the sum of the RMS currents corresponding to the 15-min or 30-min maximum demand in each of the twelve previous months, divided by 12. If twelve months are not available (short service life), sum the maximum 15/30-min apparent-power demand over the months available and divide by the number of months. For a proposed new installation, base IL on the projected 15/30-min maximum monthly apparent-power demand over the year following operation of the harmonic-producing loads on the service application. This explicit treatment removes the 2014 ambiguity that left IL undefined when twelve months of data did not exist.
TDD = iTHD × ( Imeas / IL ) where Imeas ≈ fundamental current during the testThis single relationship is the engine behind most compliance disputes: the same physical emission (a fixed iTHD in amps) produces a very different TDD depending on the IL you divide by — and IL is supplied externally, not measured during the test.
From the grid source to the PCC — and where the analyzer sits
A typical 33 kV industrial connection. The limits apply at the PCC, and that is exactly where a Class A power-quality analyzer is installed — tapped through the metering CTs and VTs, alongside the DISCOM energy meter.
Voltage, current, and what 2022 changed
Voltage limits are unchanged from 2014. The current table gained two real changes — even-order harmonics and inverter-based resources — and the column structure was adjusted to carry the even-harmonic rule.
Table 1 — Voltage distortion limits (at the PCC)
| Bus voltage V at PCC | Individual harmonic (%) | THD (%) |
|---|---|---|
| V ≤ 1.0 kV | 5.0 | 8.0 |
| 1 kV < V ≤ 69 kV | 3.0 | 5.0 |
| 69 kV < V ≤ 161 kV | 1.5 | 2.5 |
| 161 kV < V | 1.0 | 1.5 |
Table 2 — Current distortion limits, 120 V through 69 kV (the band that applies to 11 / 22 / 33 kV consumers)
The highlighted row tracks the §06 calculator below — change Isc or IL there and watch which limit band applies.
| Isc/IL (SCR) | 2 ≤ h < 11 a | 11 ≤ h < 17 | 17 ≤ h < 23 | 23 ≤ h < 35 | 35 ≤ h ≤ 50 | TDD |
|---|---|---|---|---|---|---|
| < 20 | 4.0 | 2.0 | 1.5 | 0.6 | 0.3 | 5.0 |
| 20 – 50 | 7.0 | 3.5 | 2.5 | 1.0 | 0.5 | 8.0 |
| 50 – 100 | 10.0 | 4.5 | 4.0 | 1.5 | 0.7 | 12.0 |
| 100 – 1000 | 12.0 | 5.5 | 5.0 | 2.0 | 1.0 | 15.0 |
| > 1000 | 15.0 | 7.0 | 6.0 | 2.5 | 1.4 | 20.0 |
a 2022 even-harmonic rule: for h ≤ 6, even harmonics are limited to 50% of the value shown; even harmonics of order h > 6 are limited to 100% of the value shown. The band header begins at h = 2 (2014 began at h = 3). Tables 3 (69–161 kV) and 4 (> 161 kV) carry the same structure with progressively tighter values — for Indian 11/22/33 kV industrial consumers, Table 2 is the relevant one.
Change 1 — Even-order harmonics relaxed
| Even harmonic order | IEEE 519-2014 | IEEE 519-2022 |
|---|---|---|
| h ≤ 6 (i.e. 2, 4, 6) | 25% of the odd limit | 50% of the limit (relaxed ×2) |
| h > 6 (i.e. 8, 10 …) | 25% of the odd limit | 100% of the limit (relaxed ×4) |
The change acknowledges that inverter-based resources, active front-end drives, and active filters can produce some even-order content that the stricter 2014 limits flagged unnecessarily.
Change 2 — Inverter-based resources are now in scope
In Clause 1.2 (Purpose) the word "passive" was removed: any user equipment that alters system impedance and raises voltage distortion — including inverter-based resources (IBR) and distributed energy resources (DER) — is now within the standard's purview. Per Clause 5.2, where IBR/DER generation (e.g. rooftop or captive solar) exceeds 10% of the annual average demand, the current-harmonic limits of IEEE 1547 or IEEE 2800 apply at the PCC instead of Table 2; at or below 10%, the Table 2 limits apply. The 2014 edition gave no specific treatment for grid-interactive equipment.
A full week, business-as-usual, on a Class A instrument
519 compliance is statistical, not a single reading. A meter "showing 14% TDD right now" tells you nothing about compliance until it is processed over the assessment window against the percentile rules.
Instrument and basis
- Power-quality analyzer compliant with IEC 61000-4-30:2015 Class A (Edition 3) — Edition 3 cancels and replaces the 2008 Edition 2, so a legally defensible measurement requires an Ed.3 Class A instrument.
- Under IEC 61000-4-30 Class A, every quantity is built from a 200 ms basic measurement (10 cycles at 50 Hz / 12 cycles at 60 Hz), which is then aggregated up to the 3-second very-short-time (VST) interval and the 10-minute short-time (ST) interval; harmonics are measured to IEC 61000-4-7.
- The 3-second interval is mandatory for IEEE 519-2022. The daily 99th-percentile limit is evaluated on 3-second VST values, so an instrument that does not produce the 3-second aggregation cannot generate that statistic and must not be used for an IEEE 519-2022 assessment.
- Measurement located at the PCC, captured under normal, business-as-usual facility operation.
The statistical current limits (a minimum one-week window)
- Daily 99th percentile VST (3-second) harmonic currents should be less than 2.0× the Table 2 values.
- Weekly 99th percentile ST (10-minute) harmonic currents should be less than 1.5× the Table 2 values.
- Weekly 95th percentile ST (10-minute) harmonic currents should be less than (≤) the Table 2 values.
This three-tier rule allows brief peaks while holding the sustained emission to the table limit. A few-hour snapshot cannot produce any of these percentiles and is therefore not a valid 519 assessment — only a continuous campaign over the assessment week can.
IEEE 519 is satisfied or violated only at the PCC — never at an individual drive, panel, or the LV main.
The compliance calculator — from fault level to verdict
Enter the system data and the tool first derives the available fault current Isc at the PCC (impedances combined per IS 13234 / IEC 60909). That Isc then drives the IEEE 519-2022 SCR band and the PASS/FAIL on your measured emission. Change only IL, or only the assumed fault level, and the verdict moves while the physical harmonics stay put.
Source and transformer modelled as predominantly reactive (R ≪ X at HV); line R and X are entered explicitly. Voltage factor c = 1.0 — IS 13234 / IEC 60909 apply c = 1.10 for the maximum short-circuit current above 1 kV, which would raise Isc by ~10%.
Preset A reproduces a documented educational-institution case: the same system gives Isc ≈ 1.5 kA, and iTHD 9.8% at 33 kV becomes TDD 12.45% against an 8% limit (FAIL) when IL is the pandemic-suppressed 32.96 A — yet the identical emission gives 7.97% (PASS) at the correct IL of 51.5 A. Preset B holds a 9% emission fixed and changes only the assumed upstream fault level: the true 5000 MVA source puts SCR just above 50 (12% limit, PASS), while an understated 1500 MVA drops SCR just below 50 (8% limit, FAIL) — the verdict turns on the fault level at a Table-2 band boundary. Note: this single-window check illustrates the Isc/IL effect — a field verdict applies the §05 daily/weekly percentile rules over a full week.
The recurring misreadings of 519
Each of these is heard regularly in compliance discussions. Each is wrong under 519-2022, and the correct reading is short.
Compliance is decided by TDD, not iTHD. iTHD is referenced to the fundamental and is only used to compute TDD (referenced to IL). A high iTHD at light load can be fully compliant.
At light load iTHD rises while the harmonic amperes fall — and it is the ampere value that heats equipment and distorts voltage. Treating a high light-load iTHD leads to oversized, sometimes harmful, filters.
No such requirement exists in 519. The standard's compliance metric (TDD) already accounts for facility size through IL. Forcing artificial loading only invites gaming.
The limits apply only at the PCC. Applying them to individual equipment is a deliberate engineering choice some make for design margin — it is not what the standard requires for compliance.
Compliance is statistical over a minimum one-week window, evaluated against daily/weekly 99th and 95th percentile rules. A snapshot cannot generate those percentiles.
The PCC is upstream of the installation — on the HV / incoming side for a consumer with its own transformer. Measuring at the LV main misplaces the boundary and the limits.
- Isc is the actual available fault current at the PCC, computed from source MVA and series impedance. An assumed Isc mis-selects the SCR band and therefore the TDD limit — as Preset B above shows.
- In the field, TNPDCL has applied IEC 60076-5 — the transformer short-circuit-withstand standard — to this determination (Memo No. CE/Coml/SE/EE3/AEE2/F-SC-HARMONICS/D./FLM/CM/380/2022, dt. 02-01-2023). That standard concerns a transformer's ability to withstand short circuit, not the available fault current at the PCC; using it to set Isc or the SCR band is an unconventional and incorrect practice under IEEE 519-2022, which requires the actual computed fault level (IS 13234 / IEC 60909).
That was 2014. Under 2022, even harmonics h≤6 are limited to 50% and even harmonics h>6 to 100% of the table value.
2022 brings IBR/DER into scope. Above 10% of annual average demand, IEEE 1547 / 2800 current limits apply at the PCC instead of Table 2.
Both pass the paperwork while the grid still receives the harmonics in normal operation. Neither is mitigation; both waste energy and defeat the standard's purpose.
Field practices that produce wrong verdicts
These are the practical errors that put a compliant consumer in default — or let a non-compliant one through. Most arise from the external numbers (Isc, IL) and from shortcuts on instrument and window.
Fault levelWrong Isc → wrong SCR → wrong limit band
Using an assumed, rule-of-thumb, or sanctioned-demand-derived short-circuit current instead of the actual available Isc at the PCC. Because the TDD limit is selected by SCR = Isc/IL, an understated Isc tightens the band and can fail a compliant consumer; an overstated Isc relaxes it and lets emission through.
Demand currentWrong IL → TDD inflated or deflated
Common forms: using sanctioned demand instead of the 12-month maximum-demand average; carrying forward a pandemic- or shutdown-suppressed IL; not using the 15/30-min demand interval; or ignoring the partial-year and new-installation rules. When IL is well below the operating current, TDD inflates (consumer penalised); when far above, TDD deflates (utility disadvantaged) — though the actual emission is identical.
WindowShort measurement instead of a full week
Taking a few-hour or single-visit snapshot and reading instantaneous TDD off the meter. This cannot generate the daily/weekly percentiles the standard requires, and it is unrepresentative of normal operation.
InstrumentNon-Class-A meter used for a statutory check
An ordinary energy meter or basic analyzer that is not IEC 61000-4-30 Class A (Ed.3) does not aggregate or compute percentiles to the standard, leaving any result disputable.
MetriciTHD compared directly to the TDD limit
Reading iTHD and testing it against the Table 2 TDD value. iTHD is referenced to the fundamental, not IL, so this is comparing two different quantities.
LocationMeasuring at the wrong point
Connecting at the LV main or at individual loads rather than at the contractual PCC misplaces the boundary; distortion is always higher near the loads than at the upstream PCC.
Temporary filters and inflated demand — passing the test, failing the grid
A weekly campaign captures only a snapshot of the year. This is precisely why power quality calls for continuous monitoring, and why a permanently installed Class A (IEC 61000-4-30 Ed.3) instrument at the consumer's PCC is the sound basis for compliance.
IEEE 519-2022 and the CEA Regulations rest on shared responsibility: the consumer limits the harmonic current injected at the PCC, and the licensee keeps the supply voltage within limits. In a sound arrangement a Class A instrument is installed at each consumer's PCC, the consumer shares its current-distortion data with the DISCOM, and the DISCOM in turn shares the voltage-quality data with the consumer.
With respect, we observe that this reciprocal duty has not yet been taken up by the DISCOM. In practice the obligation is placed largely on the consumer, and compliance is judged from a short, ~7-day temporary measurement. Where that measurement is also based on an incorrect short-circuit level — as in the TNPDCL practice noted in §07 — the verdict no longer reflects how the facility actually behaves across the year. That gap is what opens the door to the two short-window possibilities described below.
The window can be gamed; the grid cannot.
Because the assessment runs over roughly a week, there are two short-window possibilities that could, in principle, produce a "PASS" that does not reflect how the facility behaves across the year:
1. A temporary filter limited to the measurement window. There is a possibility that a harmonic filter is fitted only for the ~7–10 day campaign and removed afterwards. The report would then show compliance, while for the remaining ~355 days the unfiltered harmonics would flow into the grid as before — and neighbouring users, including those under no harmonic obligation, would still receive the distorted voltage.
2. A higher demand held during the window. Equally, there is a possibility that a high maximum demand is held through the campaign, lifting the 12-month IL so that the operating current sits below IL and the TDD deflates. This would waste energy and surrender demand headroom for no real benefit, and would in any case be temporary — next year's IL resets the calculation.
Neither possibility is mitigation. The standard's intent is to limit what the facility injects in normal operation; a result that depends on temporary equipment or contrived loading would not represent that — a properly designed permanent filter, continuous monitoring, and sharing the data with the DISCOM remain the only genuine remedy. On the other hand, the DISCOM should equally install power-quality meters at its substation (SS) bus and share the voltage-quality data with the public.
Using an assumed Isc instead of the actual computed fault level pulls in two opposite directions, and neither serves the grid:
1. An understated fault level penalises genuine consumers. A low assumed Isc lowers the SCR and assigns a TDD limit stricter than the connection actually warrants. Compliant consumers are then recorded as non-compliant and asked to mitigate emissions that were within a fair limit — in effect, victimised by a number.
2. An overstated fault level on a genuinely weak grid permits over-injection. Where the real fault level is low but a higher value is assumed, the SCR and the TDD limit are inflated, and consumers are allowed to inject more harmonics than a weak network can absorb. The grid is left polluted, neighbouring users — including those under no harmonic obligation — bear distorted voltage, and the added harmonic heating of transformers and lines becomes a recurring technical and revenue loss that the licensee ultimately carries.
We submit, in good faith and as a matter of engineering record, that the present unconventional practice — short windows, an incorrect Isc, and the absence of continuous Class A monitoring — has the unintended effect of permitting, and even encouraging, pollution of the power system, over and above the avoidable losses it creates. A single, mutually agreed method of determining Isc and IL, supported by permanent Class A metering and two-way data sharing, would set this right for both the consumer and the DISCOM.
Where a facility has a genuinely abnormal IL (post-pandemic recovery, major expansion, prolonged supply disruption, or efficiency works that lowered load), a fair assessment should run a continuous full-month measurement and treat the maximum-week current — the week in which the highest absolute harmonic generation occurs — as the basis for IL. This protects both the grid and the consumer from a distorted verdict.
This is Foretec's recommended interpretation and a proposed amendment — it is not present in the current IEEE 519-2022 text, which defines IL on the 12-month demand basis of Clause 3.1.
What a defensible 519-2022 assessment looks like
A measurement is only as sound as the two external numbers and the window behind it. The checklist below is what removes discretion and dispute.
- Isc computed from actual source MVA + series impedance.
- IL per Clause 3.1 (15/30-min, 12-month / partial / projected).
- SCR band and TDD limit read from Table 2/3/4.
- Class A Ed.3 instrument; harmonics per IEC 61000-4-7.
- Minimum one-week window, business-as-usual operation.
- 99th/95th percentile evaluation, not snapshots.
- TDD (not iTHD) used as the compliance metric.
- Even-harmonic 2022 rule applied (50% / 100%).
- IBR/DER >10% routed to IEEE 1547 / 2800.
- Automated report, no manual recalculation.
- Any permanent filter properly designed — not a test-week loaner.
CEA / TNERC should adopt a single, mutually agreed protocol for determining IL and Isc at the PCC, so that neither the licensee nor the consumer can apply discretionary values. A standardised determination of these two numbers would close the majority of harmonic-compliance disputes at source — before any meter is connected.
Standards and regulatory sources
- IEEE Std 519-2022, IEEE Standard for Harmonic Control in Electric Power Systems — IEEE Standards Association. standards.ieee.org/ieee/519/10677 · IEEE Xplore: document/9926021
- Central Electricity Authority (Technical Standards for Connectivity to the Grid) (Amendment) Regulations, 2019. cea.nic.in — Connectivity Amendment 2019
- Tamil Nadu Electricity Regulatory Commission — Amendments to the Tamil Nadu Electricity Supply Code, Notification No. TNERC/SC/7-47 dated 29.03.2022 (Lr. No. TNERC/Legal/1030/2022). tnerc.tn.gov.in
- IEC 61000-4-30, Testing and measurement techniques — Power quality measurement methods (Class A, Ed. 3). webstore.iec.ch/publication/68642
- IEC 61000-4-7, General guide on harmonics and interharmonics measurement and instrumentation. webstore.iec.ch/publication/4228
- Ravichandran K., Foretec Electric & A. Subbaiya (2022), Realities in Harmonics Mitigation as per IEEE 519 — Way Forward, Series 1 — field-study white paper on the indexing of harmonic emission through Total Demand Distortion and the decisive role of maximum demand current (IL). Foretec Electric India Pvt Ltd.
- Foretec Electric India Pvt Ltd — application note, IEEE 519-2022: What Has Changed from IEEE 519-2014. foretecelectric.com