One snapshot from a hand-held meter neither passes nor fails IEEE 519.
Harmonic distortion moves with the load. A fan wall at night speed and a chiller staging on give different readings at the same busbar within the hour. IEEE 519-2022 deals with that by fixing what is measured and how the results are reduced before they meet the limits.
The limits apply at the point of common coupling, where the site meets the utility, not to individual pieces of equipment, so a reading at the fan wall's own switchboard proves nothing either way.
The analyser aggregates every measurement over 3 seconds, the very short time value, and over 10 minutes, the short time value, and keeps both series for a week.
What meets the limit is a percentile of those series, not the worst reading: the daily 99th percentile of the 3 s values, and the weekly 95th and 99th percentiles of the 10 min values. For current, the daily 99th percentile is allowed 2.0 times the table value, the weekly 99th 1.5 times, and the weekly 95th the table value itself. A spike that lasted a minute sits inside the top 1% of its day and is not, by itself, a failure.
The current limits are given as TDD, not the THDi on the meter. Our own design target is in the meter's terms, THDi ≤ 5% at the PCC, and it is a target, not an IEEE 519 limit.
So a compliance check is a week with a power quality analyser at the PCC, during normal production.
Topics




