The limits apply at the point of common coupling, not at the drive — and that distinction decides your entire mitigation budget.
Nearly every harmonic-related RFQ we see contains some version of the line "must meet IEEE 519." Taken literally that requirement is incomplete, and the ambiguity routinely costs buyers either a failed commissioning test or tens of thousands of dollars of filtering they never needed.
IEEE 519-2022, Recommended Practice and Requirements for Harmonic Control in Electric Power Systems, is a shared-responsibility document. It sets limits on the harmonic current a customer injects into the utility system, and limits on the voltage distortion the utility delivers. Both are defined at one specific place.
The point of common coupling (PCC) is the electrical boundary between you and the utility — typically the service entrance or the primary of the utility transformer, wherever another customer could also be served. It is not the drive terminals, the motor control center, or the branch circuit feeding the non-linear load.
This matters enormously. A 100 HP variable frequency drive with no mitigation might draw 35–40% current distortion at its own terminals. Measured at a PCC that also carries several megawatts of linear load, that same drive may contribute well under 5% total demand distortion. The drive has not changed. The measurement point has.
The second common error is specifying THD where the standard uses TDD. They are different denominators:
At light load, THD can read alarmingly high while the actual harmonic amperes flowing are trivial. A drive idling at 10% load might show 60% THD and still be entirely compliant, because TDD normalises against full demand. Measuring THD at light load and declaring a failure is one of the most common false alarms in commissioning.
IEEE 519 does not impose one number on everybody. The allowable TDD scales with the stiffness of the supply relative to your load — the ratio of available short-circuit current (Isc) at the PCC to maximum demand load current (IL). A stiff source tolerates more harmonic current because the resulting voltage distortion is smaller.
| Isc / IL | h < 11 | 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% |
Current distortion limits for systems rated 120 V through 69 kV, expressed as a percentage of IL. Even harmonics are limited to 25% of the odd harmonic limits above them.
The practical reading: an installation on a stiff service with a modest drive load may have a 15% or 20% TDD allowance, which a set of 5% line reactors will comfortably meet. An installation on a weak service — a remote site, a generator, an overloaded transformer — may be held to 5.0%, which usually requires tuned passive filtering or an active solution.
For systems at the PCC rated 1 kV and below, IEEE 519-2022 limits individual harmonic voltage to 5.0% and total harmonic voltage distortion to 8.0%. Above 1 kV through 69 kV, the limits tighten to 3.0% individual and 5.0% total. These are the utility's obligation, but they matter to you: if incoming voltage distortion is already near the limit before your equipment energises, your available headroom is smaller than the current tables suggest.
Once the limit is known, the options fall into a rough order of cost and effectiveness:
The right choice is set almost entirely by the TDD limit the Isc/IL ratio hands you, and by whether the load is steady or varying. That is why the short-circuit data is the first question, not an afterthought.
At the point of common coupling — normally the service entrance or utility transformer primary. Limits do not apply at individual drive terminals, and specifying them there usually demands far more mitigation than the standard requires.
THD expresses harmonic content against the fundamental current present at the time of measurement. TDD expresses it against maximum demand load current. IEEE 519 uses TDD, which is why a lightly loaded drive can show very high THD while remaining fully compliant.
The allowable TDD is set by the ratio of short-circuit current to maximum demand load current. That ratio can move the limit from 5% to 20%, which is the difference between needing a tuned filter and needing only line reactors.
Sometimes. Line reactors typically bring current distortion into the 35 to 40 percent range at the drive, which can be well under the PCC limit on a stiff service with modest drive load. On a weak service held to 5% TDD, they will not be sufficient on their own.
Yes, materially. A generator is a much weaker source than the utility, so the Isc/IL ratio falls and the allowable TDD tightens. Passive filters also present leading power factor at light load, which can destabilise generator voltage regulators. Generator operation must be stated before filter selection.
Data center distribution and grid-scale battery storage both present harmonic and thermal problems that traditional plant experience handles badly. The differences are specific and they change the component selection.
The passive-versus-active decision is rarely about performance at full load, where both work. It is about part load, generator operation, future expansion, and whether your load profile is actually as steady as the calculation assumed.
NEMA 1, 3R, 4X and 12 are not a simple ladder of increasing protection. Each addresses different hazards, and the wrong choice either wastes money or fails in service — often through heat rather than ingress.