Undervoltage and overvoltage protection (ANSI 27/59) — why a generator or motor also needs protection against its own terminal voltage

The guide on reverse power protection (ANSI 32) and the [guide on directional overcurrent protection (ANSI 67)](/guides/nen-3140/richtingsafhankelijke-overstroombeveiliging-67-richtingsrelais) both cover protection functions that respond to the current through a measuring point. This article covers a different category of protection, one that responds directly to the voltage itself: undervoltage protection (ANSI code 27) and overvoltage protection (ANSI code 59).

Why voltage protection has a different function than overcurrent protection

Overcurrent protection guards against an excessive current, usually caused by a fault or overload elsewhere in the network. Voltage protection guards against something different: a terminal voltage that, without there necessarily being an overcurrent, falls outside the range within which connected equipment can operate safely and reliably. A generator running in parallel with the grid, a large motor, or a busbar feeding several sensitive loads can be damaged or trip unintentionally due to a voltage dip or surge, even if the current at that moment stays within normal limits.

Undervoltage protection (27): what it protects against

A sustained excessively low terminal voltage can cause several problems:

For motors, the torque at a given load decreases with the square of the voltage, so a motor can stall under a significant undervoltage and subsequently draw a sharply increased current — see also the guide on motor protection.
For a generator, a sustained undervoltage can indicate a problem with excitation, or a heavy load elsewhere in the network pulling the voltage down.
A prolonged undervoltage can also indicate a (near-)lost grid connection, where the voltage sags before the current actually disappears.

An undervoltage relay (27) monitors the terminal voltage and trips (or initiates a transfer to another source) once the voltage stays below the set threshold for the set time delay — generally in the order of 80% to 90% of nominal voltage, with a time delay of a few seconds to about ten seconds, depending on the application.

Overvoltage protection (59): what it protects against

A sustained excessively high terminal voltage stresses the insulation of connected equipment above its design value and, depending on the cause, can point to:

An incorrectly set or failing voltage regulation of a generator or of a tap-changing regulating transformer (OLTC).
The Ferranti effect or a suddenly lost load on a long, lightly loaded circuit.
A ferroresonance condition, see the guide on ferroresonance in a voltage transformer, where the voltage can rise well above normal values.

An overvoltage relay (59) monitors the terminal voltage on the upper side and trips once the voltage stays above the set threshold for the set time — generally in the order of 110% to 130% of nominal voltage. Overvoltage protection is often implemented with a two-stage setting: a moderate overvoltage with a time delay of a few seconds, and an extreme overvoltage with a much shorter, near-instantaneous trip to protect equipment insulation against a rapidly rising, potentially damaging voltage.

Why a time delay is needed here too

Just as with reverse power protection (see the guide on ANSI 32), a voltage relay is not applied instantaneously: a brief voltage dip while a large motor is starting, or a short voltage spike during a switching operation elsewhere in the network, must not immediately lead to a trip. The time delay is tuned to distinguish between such a harmless, transient event and a genuinely sustained under- or overvoltage that warrants intervention.

Note: the exact thresholds and time delays for 27/59 follow from the system study of the specific generator, motor, or switchgear and from the requirements of the grid operator (for example when connecting distributed generation); this article covers the principle, not a ready-made setting table for every application.

Practical relevance

When commissioning a generator or a busbar feeding critical loads, it is important to verify that both the undervoltage and the overvoltage protection functions are actually configured and functionally tested — the absence of either function leaves connected equipment unprotected against exactly the part of the voltage range that function is meant to cover.

Common mistakes

Applying only overcurrent protection and skipping voltage protection — a voltage problem without overcurrent (for example a failing voltage regulator) then goes unnoticed until actual damage occurs.
Setting a single, overly tight threshold for both normal starting dips and a genuine undervoltage fault — this leads to nuisance tripping when, for example, a large motor starts.
Not implementing overvoltage protection with a two-stage setting — a single, moderate time delay protects insulation insufficiently against a rapidly rising, extreme overvoltage.
Not reviewing the 27/59 settings after a change to the network configuration or after connecting new distributed generation — the normal voltage bandwidth of the network can shift as a result, making previously correct settings unsuitable.