The governor's droop setting is a single number — typically expressed as a percentage. It sits in the governor's parameter table next to dozens of other settings. It is almost never the headline parameter in a plant's marketing material.
It is also the parameter that, more than any other, determines how that plant behaves during a grid frequency excursion — and therefore how the plant participates in maintaining the stability of the entire interconnected power system.
Why 3% to 7.5% is the codified band
Most grid codes specify an acceptable droop band for synchronous generation participating in the wholesale market. The codified band is typically 3% to 7.5% — that is the value in Mexico's Código de Red 2.0, and the same range (sometimes 2% to 8%) appears in NERC's reliability standards, ENTSO-E's European framework, Chile's NTSyCS, and most other regulated systems.
Inside that band, the plant operator and the regulator have legitimate engineering room to negotiate the specific setting. The choice depends on prime mover, fuel system response, control loop tuning, and the unit's economic operating point. There is no single correct answer inside the band; there are wrong answers outside it.
How the test is actually run
The unit is operated stably at a known power output on automatic governor control. A controlled frequency setpoint change is introduced at the governor reference. The unit's power output response is recorded. The ratio of the frequency change to the resulting power change (corrected for nominal values) yields the measured droop.
Why this test is the regulator's leverage
Droop is one of the few plant parameters the regulator can validate experimentally from outside the plant. The system operator cannot easily verify that a plant's claimed heat rate is what's printed in the dossier. They can verify droop because they can perturb the governor's frequency reference and watch what the unit does. That makes the droop test one of the most rigorously reviewed items in the field campaign.