Automatic Transfer Switches: Types, Transitions and Application

What an ATS does and how it is built

An automatic transfer switch (ATS) listed to UL 1008 monitors the normal source, signals the alternate source to start when the normal source fails, transfers the load once the alternate source is within acceptable voltage and frequency, and transfers back when the normal source has been stable for a set time. Ratings run from tens of amperes to 4,000 A and above, in three-pole and four-pole versions, the latter with a switched or overlapping neutral for separately derived sources.

Three construction types are common. Contactor-type switches use a double-throw mechanism with mechanically interlocked contacts and carry no overcurrent protection of their own. Breaker-type switches pair two interlocked molded-case or power circuit breakers with motor operators and can include overcurrent protection. Every ATS also carries a withstand and closing rating (WCR) that must be coordinated with the upstream overcurrent device, either by specific breaker type, by any breaker up to a rating, or by a time-based rating.

Open transition

Open transition is break-before-make. The switch opens the connection to the normal source, passes through a neutral position and closes to the alternate source, and the load sees an interruption of a few cycles to a few seconds depending on the mechanism and the programmed delays. The two sources are never connected together, so there is no paralleling and no utility involvement.

It is the simplest and most common transition and is appropriate for lighting, receptacles, HVAC and most building loads. For motor loads the transfer can be supervised by an in-phase monitor, which waits for the two sources to drift into phase before transferring so the motor is not reconnected against its own residual voltage.

Closed transition

Closed transition is make-before-break. When both sources are present and within a voltage, frequency and phase-angle window, the switch closes to the alternate source before opening from the normal source, paralleling them for a brief interval, typically under 100 milliseconds. The load sees no interruption, which is why closed transition is used for planned tests, retransfer to the normal source and load management.

Because the two sources are momentarily connected, paralleling with the utility requires utility approval and often protective relaying agreed with the utility, and the fault current during the parallel must be within the ratings of the equipment. If the sources do not synchronize within a programmed timeout, the controller falls back to open transition. On loss of the normal source there is nothing to parallel with, so every closed-transition switch performs an open transition in an actual outage.

Delayed transition

Delayed transition is an open transition with a deliberate, programmable pause in the neutral position, commonly from half a second to several seconds. The pause lets motor residual voltage decay, lets transformer magnetizing flux settle and gives UPS rectifiers and variable-frequency drives a clean restart rather than an abrupt source change with a phase jump.

It is specified for large motor loads, drive-heavy process loads, transformer-fed distribution and UPS inputs. Many delayed-transition switches also provide pre-transfer signal contacts that warn elevator controllers, chillers and drives a few seconds before the transfer so they can shed or prepare.

Bypass-isolation and service-entrance rated switches

A bypass-isolation ATS adds a manually operated bypass switch in parallel with the automatic switch. The automatic section can be placed in test or isolated from both sources and the load, inspected, tested and even removed while the bypass carries the load. NFPA 110 recommends bypass-isolation where an emergency power supply system must be maintained without interrupting the load, and hospital and data center specifications routinely require it.

A service-entrance rated ATS incorporates the service disconnecting means and overcurrent protection, with the neutral-to-ground bond made at the switch. Where the service is 1,000 A or larger on a solidly grounded wye system over 150 V to ground, the NEC requires ground-fault protection of equipment at the service disconnect, and the ATS must accommodate it.

Timing, codes and testing

The NEC classifies alternate-power systems into Article 700 emergency systems, which must have power available within 10 seconds, Article 701 legally required standby systems, which allow 60 seconds, and Article 702 optional standby systems, which have no time requirement. NFPA 110 classifies the emergency power supply system by type (time to restore power), class (run duration) and level (criticality) and governs installation, testing and maintenance of the system the ATS belongs to.

ATS controller settings implement this. A short engine-start delay, often 0.5 to 3 seconds, prevents starting on a momentary dip. A normal-to-emergency delay lets the generator stabilize. A retransfer delay of several minutes confirms the normal source is stable before returning. An engine cooldown timer runs the generator unloaded after retransfer.

The controller also provides an exerciser schedule and test switch, and NFPA 110 requires periodic testing of the system under load. Report position, source availability and alarms to the building management or SCADA system so an unnoticed failed transfer is not discovered during the next outage.