Flickermeter — Pst/Plt measurement of voltage flicker per IEC 61000-4-15

The guide on flicker and voltage fluctuations covers why rapidly varying loads (welding equipment, large motors with varying load, arc furnaces) can cause objectionable light flicker. This article covers the instrument and the standardised measurement method used to objectively quantify that nuisance: the flickermeter per IEC 61000-4-15.

Why an ordinary voltage measurement doesn't capture flicker nuisance

An ordinary RMS voltage measurement records a voltage dip or a slow trend, but says little about how annoying a rapidly fluctuating voltage is perceived to be by the human eye. The eye's sensitivity to light flicker is strongly frequency-dependent — flicker occurring a few times per second is perceived as far more annoying than very slow or very fast fluctuations. IEC 61000-4-15 describes a flickermeter that processes the signal through a standardised model of a 60 W incandescent lamp and the human eye-brain system, so the outcome yields an objective measure of actually perceived nuisance, rather than a raw voltage fluctuation in volts or percent.

Pst: the short-term indicator

Pst (short-term flicker severity) is determined over a 10-minute period and reflects the degree of flicker nuisance statistically observed during that period. A Pst value of 1.0 corresponds to the level at which an average observer just perceives the light flicker as irritating — this is the reference against which limit values are assessed.

Plt: the long-term indicator

Plt (long-term flicker severity) is calculated from twelve consecutive Pst values over a 2-hour period, using a cumulative formula that weighs both short, severe flicker peaks and longer, milder flicker. Plt is representative of the nuisance caused by a repeating or varying load pattern over a longer period, such as a production process with varying load cycles over a shift.

Note: EN 50160 requires that, at the connection point under normal operating conditions, Plt ≤ 1 for 95% of the time in each one-week period. This limit applies to the long-term indicator; a brief Pst peak above 1.0 is not by definition a violation as long as the Plt over the full 2-hour period stays within the limit. (The value Plt ≤ 0.65 belongs to a different type of standard — the equipment emission limit under EN 61000-3-3/3-11 — not the network connection characteristic of EN 50160.)

Practical deployment

A flickermeter, like the power-quality datalogger, is typically connected for an extended period at the point where the complaint occurs or where the grid operator assesses the connection condition. Some practical points to note:

Measurement duration: because Plt is built from twelve Pst values, a continuous measurement of at least several hours is needed to obtain a single reliable Plt value; a measurement lasting several days to a week is typical for a representative picture of a daily or weekly load pattern.
Measurement point: flicker must be measured at the point where the nuisance actually occurs or where the grid connection condition applies — a measurement behind a private transformer does not automatically represent the flicker at the grid operator's connection point.
Correlation with load events: as with the power-quality datalogger, time synchronisation is essential to link a Pst peak to a specific process event (a welding machine starting, a large motor starting, an arc furnace cycle) rather than merely reporting an isolated figure.
Distinguishing the cause: an elevated Plt can be caused by the facility's own installation (its own varying load) or by another customer on the same network — the flickermeter itself does not make that distinction; additional analysis of the load pattern is needed for that.

Practical relevance

For a complaint about light flicker, or when substantiating a connection condition to the grid operator for a new, strongly varying load (for example a welding robot or a large variable-frequency drive with a cyclic load pattern), a flickermeter measurement per IEC 61000-4-15 is the appropriate, objectifiable instrument — in contrast to a subjective assessment of "the light flickers here sometimes".

Common mistakes

Flagging a brief Pst peak as a limit violation on its own, while the relevant limit value (Plt) is assessed over a full 2-hour period.
Measuring for too short a period to obtain a representative Plt, causing an incidental, non-representative period to be wrongly labelled as a structural problem, or vice versa.
Measuring at a non-representative point (for example behind a private transformer) instead of at the actual connection point to which the standard applies.
Not maintaining time synchronisation with the process sequence, which prevents a measured Pst peak from being traced back to the load event that caused it.