Reverse power protection (ANSI 32) — why a generator that keeps turning is no proof that all is well

The guide on paralleling and synchronizing standby generators covers how a generator is matched in frequency, voltage, and phase angle before being connected to the grid or to another generator. This article covers what can happen after that connection has been made: the generator's drive (engine, turbine) can fail while the generator itself keeps turning quietly through its electrical connection — and why this requires a specific protection function, reverse power protection, denoted by the ANSI code 32.

Why a generator without a drive does not simply stop

A synchronous generator connected in parallel with the grid (or with another, healthy generator) is forced by that grid to keep turning exactly at grid frequency, regardless of what happens to its own drive — as long as the magnetic coupling between rotor and stator is not broken. If the drive engine or turbine fails (fuel shortage, a mechanical fault, a lost combustion process), the generator no longer delivers mechanical power to the shaft, but, forced by the grid, continues to turn at synchronous speed. The grid then supplies the power needed to cover exactly the friction, windage, and other losses of the no-longer-driven shaft: at that point the generator functions, electrically, as a synchronous motor driven by the grid — hence the term anti-motoring protection for this function.

Why this can be damaging, and why that differs by drive type

A generator running as a motor draws only a small percentage of its rated power from the grid (just enough to cover its own losses), but the problem usually is not the electrical load itself: it is the drive. For an internal combustion engine or gas turbine being carried along as a load without fuel supply or active combustion, the shaft and bearing load, lubrication, or the thermal condition of the drive can quickly become unfavorable; for a steam turbine, the absence of steam flow without adequate cooling of the last stages of blading can lead to overheating of those blades within minutes. The tolerable time a specific drive may run as a motor before damage occurs therefore differs considerably by drive type — from a few seconds to a few minutes — and partly determines how quickly the reverse power protection must operate.

Working principle: monitoring power direction, not just magnitude

A reverse power relay measures the actual (active) power flowing through the generator terminals, including its direction:

Under normal operation, the generator delivers power to the grid (positive power in the conventional measuring direction).
With a failed drive, the power direction reverses: the grid now delivers a small power to the generator (negative, or "reverse", power) to keep the shaft turning.

Once the measured power shows this reversal and exceeds a set threshold — typically only a few percent (often on the order of 0.5% to 3%) of rated generator power, well below what would be needed for normal operating fluctuations — the protection operates and sends a trip command to the generator breaker. This low sensitivity is deliberate: the power needed to cover only the own losses of the undriven shaft is typically small relative to rated power, so the protection must be able to respond to a small reverse power to act in time.

Why a short time delay is still needed

Despite the low sensitivity, a reverse power relay is not applied instantaneously, but with a short, set time delay. When paralleling a generator, or during normal, brief power fluctuations from the drive itself (for example a governor that momentarily overcompensates), the measured power can briefly go negative without the drive actually having failed. A short time delay (typically on the order of a few seconds) prevents the protection from operating during such harmless, transient moments, while still remaining well within the time that most drives can withstand running as a motor without damage.

Note: the exact power threshold and time delay are tuned to the actual, manufacturer-specified losses of the specific drive in motor operation and to the maximum tolerable time in that state; this article covers the principle, not a ready-made setting for every drive type.

Practical relevance

When commissioning a standby generator that can operate in parallel with the grid or with other generators, reverse power protection must be explicitly tested by simulating a controlled reverse power condition (for example by controllably interrupting the fuel supply with the generator still connected, according to the manufacturer's test procedure) — it is not sufficient to only check that the relay is correctly configured, without having actually observed it operate in practice.

Common mistakes

Assuming a generator that keeps turning automatically means its drive is still functioning — a generator being driven as a motor by the grid can feel completely normal electrically without reverse power protection to flag it.
Applying the same reverse power threshold to very different drive types — the tolerable time and typical loss power in motor operation differ considerably between, for example, a diesel engine and a steam turbine.
Setting the time delay too short — this can cause nuisance tripping during normal, brief power fluctuations during governing or paralleling.
Not actually functionally testing reverse power protection at commissioning — a relay that has only been checked against its settings, but has never seen a real reverse power condition, can have a hidden wiring or setting error that only surfaces when a real drive failure occurs.