An automatic transfer switch acceptance test is a field protocol that proves an emergency system will actually transfer when its normal source fails, and retransfer when that source returns. It records the identification (manufacturer, model and catalog number, serial number, voltage and current rating, phase configuration, number of poles, transition type as open, closed, or delayed, the normal and alternate sources, the controller model, and the equipment designation). The inspection covers nameplate data against the drawings, condition, anchorage, alignment, grounding and clearances, cleanliness, lubrication, attached and visible warning labels, tightness of every control connection, bolted-connection tightness, a manual transfer operation, positive mechanical interlocking between the two sources, and a thermographic survey. The electrical work covers bolted-connection resistance on the normal source, alternate source, and load terminals, insulation resistance in both switch positions and across each open pole, optional control-wiring insulation resistance, contact resistance in both positions, the control device settings and operation, calibration of the relays and timers under the relay section, phase rotation, phasing and synchronization, and the timed operation of the eight emergency functions. NETA ATS-2025 section 7.22.3 lists this as the standard acceptance sequence.
A transfer switch is the single point on which an emergency system's whole value rests, and it spends its life in one position waiting for an event it may never see. That is why so much of this sequence is operational rather than electrical: the voltage- and frequency-sensing relays have to decide that the normal source has failed, the engine start contact has to close, the transfer has to wait out its delay, the interlocks and limit switches have to prevent the two sources from ever meeting, and the retransfer and engine cool-down sequence has to unwind all of it when the utility returns. Each of those is timed and recorded because a transfer that works but takes too long has already dropped the load it was protecting. The resistance measurements serve the ordinary purpose, on three sets of terminals rather than one, and the insulation-resistance readings are taken in both source positions because the switch is only ever half-tested in one.
Acceptance: after installation and before the emergency system is relied on, as part of the standard acceptance sequence, with the relays and timers calibrated under the relay section before the timed functional sequence is run. The same form carries a maintenance service type, and a transfer switch is a strong candidate for interval-based exercising, since the mechanism, the controller, and the engine start circuit all deteriorate from disuse rather than from wear. The manual transfer operation and the interlock check come before any automatic test, and the timed sequence is normally run with the generator, so the engine-generator acceptance test is coordinated with this one.
With the switch de-energized, isolated, and grounded per site safety procedures, the identification data is recorded and the inspection is worked through: condition, anchorage, alignment, grounding and clearances, cleanliness, lubrication of the moving current-carrying and sliding surfaces, the warning labels, the tightness of every control connection, and bolted connections verified with a calibrated torque wrench or by resistance measurement, followed by a manual transfer operation and a check of the positive mechanical interlocking between the normal and alternate sources. Bolted-connection resistance is measured phase by phase at the normal source, alternate source, and load terminals. Insulation resistance is measured for one minute between phases and from each phase to ground, with the switch in the normal position and again in the alternate position, and across each open pole, with the lowest reading identified and corrected to the reference temperature. Control-wiring insulation resistance is measured where the circuit can tolerate it. Contact or pole resistance is measured in both source positions. The control device settings are verified and the as-left settings recorded, the relays and timers are calibrated and set under the relay section, and phase rotation, phasing, and synchronized operation are verified as the application requires. The functional sequence then exercises the normal source voltage- and frequency-sensing relays, the engine start sequence, the time delay upon transfer, the alternate source sensing relays, the automatic transfer, the interlocks and limit switches, the time delay and retransfer on restoration of normal power, and the engine cool-down and shutdown feature, with the transfer delay, retransfer delay, and cool-down times measured.
The identification (manufacturer, model and catalog number, serial number, voltage and current rating, phase, number of poles, transition type, normal and alternate sources, controller model, and equipment designation); the result of each visual and mechanical item, including the manual transfer, the mechanical interlock check, the warning labels, and the thermographic survey; bolted-connection resistance per phase at the normal source, alternate source, and load; the insulation-resistance test voltage, the readings between phases and from each phase to ground in both switch positions, the across-open-pole readings, and the lowest reading as measured and corrected; the control-wiring test voltage, reading, and any circuits excluded as voltage-intolerant; contact or pole resistance per pole in both positions; the control device verification and the as-left settings; the relay and timer test results; the phase rotation and phasing notes; the result of each of the eight timed functions with the measured transfer delay, retransfer delay, and engine cool-down times; and the ambient temperature, relative humidity, test equipment with calibration due date, comments, and deficiencies.
Bolted-connection resistance readings are compared with similar connections and a reading well above the lowest is investigated. Insulation resistance must meet the manufacturer's published data or, in its absence, the minimum in the standard's insulation-resistance table for the switch's class, with readings below either investigated. Control-wiring insulation resistance has a published floor of its own, and where solid-state components or other devices would not survive the applied voltage, the manufacturer's recommendation replaces the test. Contact or pole resistance must not exceed the manufacturer's high-normal range, and without manufacturer data a pole deviating well above the lowest of the adjacent poles or of similar switches is investigated. Control devices must behave the way the manufacturer's published data says they should, and the relay and timer results are judged under the relay section. Phase rotation, phasing, and synchronization must agree with the system design specifications. The eight timed functions are judged against the manufacturer's data and the system design requirements together, which is the important point: the correct transfer delay is a design decision about the load, not a value the standard publishes, so the recorded times are only meaningful against the design intent. Bolt torque follows the manufacturer's data with the standard's torque table used in its absence, and thermography is judged under its own section. See the purchased NETA standard for the table values and the design documents for the timing requirements.
UltraDb references ASTM, ANSI, IEEE, and NETA standards descriptively, to identify the published methods its forms are built from. UltraDb and Entigy Solutions are not affiliated with, endorsed by, certified by, or licensed by NETA, ASTM, ANSI, IEEE, or Megger. Standard names and section numbers are the property of their respective organizations. Always test to the edition your contract specifies.