IEEE 519 is the standard most utilities, consultants and specifiers use to judge harmonic distortion. It does two things: it limits the voltage distortion the utility should deliver, and it limits the harmonic current a customer may inject. Both are measured at one place, the point of common coupling (PCC), not at each machine. Misreading that one detail is behind most arguments about whether a plant “fails IEEE 519”.
This guide explains what the standard measures, the limits in plain words, what changed in the 2022 edition, and how to check a site properly.
Who is responsible for what
IEEE 519 splits the responsibility in two:
- The customer keeps the harmonic current it draws within limits that depend on how strong the supply is at the PCC.
- The utility keeps the harmonic voltage at the PCC within limits that depend on the voltage level.
The logic is simple: harmonic current from your drives, rectifiers and UPS systems flows back through the network impedance and creates voltage distortion that your neighbours also receive. A strong supply (high short-circuit current) absorbs more harmonic current before the voltage distorts, so it is allowed more.
Where to measure: the point of common coupling
The PCC is the point where your installation connects to the network that also supplies other customers, usually the utility side of your main transformer or the main incomer. Measuring on a single drive will almost always show high current distortion, and that is not a breach of IEEE 519. The standard is about what leaves your site.
Inside a plant, the same limits are often used as a design target at internal buses. That is good engineering practice, but it is a choice, not a requirement of the standard.
Voltage limits
| Bus voltage at the PCC | Individual harmonic | THD |
|---|---|---|
| 1 kV and below | 5.0 % | 8.0 % |
| Above 1 kV up to 69 kV | 3.0 % | 5.0 % |
| Above 69 kV up to 161 kV | 1.5 % | 2.5 % |
| Above 161 kV | 1.0 % | 1.5 % |
The 2022 edition kept the voltage limits unchanged from 2014 (summary of changes).
Current limits and the short-circuit ratio
Current limits are expressed as a percentage of IL, the maximum demand load current at the PCC, and the total is expressed as TDD (total demand distortion), not THD. The right row of the table depends on the short-circuit ratio ISC/IL.
| ISC/IL | h < 11 | 11–16 | 17–22 | 23–34 | 35–50 | TDD |
|---|---|---|---|---|---|---|
| Below 20 | 4.0 | 2.0 | 1.5 | 0.6 | 0.3 | 5.0 |
| 20 to 50 | 7.0 | 3.5 | 2.5 | 1.0 | 0.5 | 8.0 |
| 50 to 100 | 10.0 | 4.5 | 4.0 | 1.5 | 0.7 | 12.0 |
| 100 to 1000 | 12.0 | 5.5 | 5.0 | 2.0 | 1.0 | 15.0 |
| Above 1000 | 15.0 | 7.0 | 6.0 | 2.5 | 1.4 | 20.0 |
Values for systems from 120 V to 69 kV, in percent of IL. Higher voltage levels have their own, tighter tables.
Why TDD and not THD? Current THD is measured against the current flowing right now. At light load, a drive can show 80 % current THD while injecting very few harmonic amps. TDD divides by the maximum demand current, so a lightly loaded site is not penalised for a percentage that means nothing. IL is normally taken from the demand history, such as the average of the monthly peak demands over 12 months.
How to get ISC. Ask the utility for the available short-circuit current at the PCC, or estimate it from the transformer rating and impedance (ISC ≈ rated current ÷ per-unit impedance) as a first approximation that ignores the upstream network.
What changed in IEEE 519-2022
The main change concerns even harmonics. The 2014 edition limited every even order to 25 % of the odd-harmonic limit in the same range. The 2022 edition relaxed this: orders 2, 4 and 6 are limited to 50 % of the table value, and even orders above 6 to 100 %. The change recognises that modern equipment such as active front-end drives, inverter-based resources and active filters can produce some even harmonics without causing problems (MIRUS application note AN003). The overall distortion budget, THD for voltage and TDD for current, did not grow.
Statistics, not single readings
Like EN 50160, IEEE 519 is assessed on statistics over time, not on one reading. Measurements are aggregated into very short (3-second) and short (10-minute) values and evaluated with daily and weekly percentiles. In practice this means:
- a short spike above the limit is not a failure on its own;
- a one-hour measurement cannot prove compliance;
- you need at least a week of data that covers your full production cycle.
Check the exact percentile rules in the standard before you write a compliance report.
How to check a site in five steps
- Find the PCC and ISC. Agree with the utility where the PCC is and what short-circuit current applies there.
- Establish IL. Use 12 months of demand data, not today’s load.
- Measure for at least a week at the PCC. Record voltage and current harmonics to the 50th order with 10-minute aggregation, plus events.
- Screen the results. Our THD calculator compares a spectrum with the IEEE 519-2022 limits, including the even-harmonic rule.
- If you are over, find the source before buying a filter. Measure the largest non-linear loads at the same time as the PCC. Read harmonics and power factor before adding capacitors, because capacitors can create a resonance that makes harmonics worse.
How Inspector helps
Inspector records harmonic amplitude and phase, THD, interharmonics and K-factor continuously on two isolated channels, so the PCC and the largest non-linear load can be measured on one instrument at the same time. A permanent Inspector V4 keeps the record running after the first week, which is what you need when harmonic levels change with production, season or a neighbour’s new load. It does not replace the formal compliance assessment set out in the standard; it gives you the data to know where you stand before that conversation.
Book a call with an Inspector engineer to plan a harmonic study, or start with the free power quality assessment.
