8.3 Automatic Transfer Switches and Emergency Power Transfer

Automatic Transfer Switches and Emergency Power Transfer

Quick Answer: An automatic transfer switch monitors the normal source and moves critical load to a standby generator or an alternate utility feed when the normal source fails, per NETA ATS/MTS Section 7.22 and NFPA 110. The distinction that drives everything: open transition breaks before it makes, producing a dead interval, while closed transition makes before it breaks, briefly paralleling two sources.

---

1. Why an ATS is a testing problem, not just a switching device

An ATS sits at the boundary between two independent power sources, and it must never connect them together unintentionally. That single requirement generates the whole test programme: mechanical and electrical interlocks that physically prevent paralleling, sensing that correctly recognizes source failure and source restoration, timing that prevents nuisance transfers, and a control sequence that starts, loads, unloads, and cools an engine.

It also matters for isolation planning. As covered in the drawings section, an ATS is a source. Work downstream of an ATS is not isolated by opening the normal supply alone — that is precisely the condition the ATS is built to respond to.

2. Transition topologies

+-----------------------------------------------------------------------------------------+
|                           ATS OPERATIONAL TRANSITION MODES                              |
|                                                                                         |
|   1. Open Transition (Break-Before-Make):                                               |
|      Normal [CLOSED] ---> [OPEN] ---> [DEAD TIME] ---> Emergency [CLOSED]               |
|      - In-Phase Transfer: transfers only within a narrow phase-angle window.            |
|      - Delayed (Programmed) Transition: pauses in a center-off neutral position to      |
|        let motor residual back-EMF decay to zero before reconnecting.                   |
|                                                                                         |
|   2. Closed Transition (Make-Before-Break):                                             |
|      Normal [CLOSED] ---> Emergency [CLOSED] (parallel, brief) ---> Normal [OPEN]       |
|      - No interruption to the load; requires both sources present and in synchronism.   |
+-----------------------------------------------------------------------------------------+

Closed transition is only available when both sources are healthy, which means it works for a planned transfer — a generator exercise or a scheduled utility outage — but cannot help during an actual utility failure, because there is nothing to parallel with. A site that specifies closed transition still gets an open-transition event when the utility genuinely fails.

3. The motor residual voltage hazard

This is the single most examinable ATS concept.

A spinning induction motor does not stop generating when its supply is removed. Rotor flux decays over several seconds, and during that decay the motor produces a residual back-EMF at its terminals — at a frequency that falls as the rotor slows, so it drifts steadily out of phase with the incoming source.

If the ATS reconnects while the residual voltage is roughly 180° out of phase with the new source, the instantaneous voltage across the motor terminals approaches the sum of the two — on the order of 2 to 3 times nominal. The resulting inrush can reach many times full-load current, and the mechanical torque transient is what actually does the damage: snapped shafts, stripped couplings and gearboxes, and displaced windings, alongside tripped breakers.

The two mitigations:

• In-phase monitor — compares the phase angle between the decaying motor voltage and the oncoming source, and permits transfer only inside a narrow alignment window. Fast, but it requires both sources present to compare.
• Delayed neutral (programmed) transition — parks the switch in a center-off position for an adjustable interval so the rotor flux decays essentially to zero before the new source is applied. Slower, and the load is dead for the duration, but it works without needing the second source for reference.

Selecting between them is an engineering decision about load tolerance for the dead interval; verifying that the selected function actually operates is a commissioning test.

4. NETA test sequence (Section 7.22)

+-----------------------------------------------------------------------------------------+
|                         ATS TESTING PROTOCOL SUMMARY MATRIX                             |
|                                                                                         |
|   [1. CONTACT RESISTANCE]        [2. INTERLOCK VERIFICATION]    [3. SENSING CALIBRATION]|
|   - DLRO 4-wire measurement.     - Mechanical AND electrical    - Verify pickup/dropout |
|   - Compare across poles.          must prevent paralleling.      on both controllers.  |
|                                                                                         |
|   [4. TIME DELAY CHECKS]         [5. ENGINE START CONTACT]      [6. FULL LOAD TRANSFER] |
|   - TDES, TDNE, TDEN, TDEC.      - Verify dry-contact closure.  - Simulate utility loss;|
|                                                                   verify full cycle.    |
+-----------------------------------------------------------------------------------------+

Visual and mechanical: enclosure and NEMA rating intact, cable terminations torqued to specification, arc chutes and barriers in place, manual operating handle functional, and both the mechanical and the electrical interlock physically verified — attempt to defeat each and confirm the other holds. An interlock that is inspected but never functionally tested has not been verified.

Contact resistance: measure across the normal-to-load and emergency-to-load main contacts with a four-wire DLRO, compare pole to pole and against the manufacturer's limit.

Insulation resistance: pole to pole and pole to ground, in both the normal and emergency positions.

Sensing calibration — the values to know:

The hysteresis between dropout and pickup is deliberate: a switch that transferred at 80 % and retransferred at 80 % would chatter continuously on a marginal supply. The separation is what makes the decision stable.

Time delays:

Time each with a stopwatch against the setting; do not accept the display value as proof.

Engine start contact: verify the dry-contact closure and its circuit continuity. A control scheme that transfers perfectly but never signals the engine to start leaves the load dead.

5. Full outage simulation

The commissioning test that proves the system rather than its parts:

1. Open the normal utility supply to simulate a genuine outage.
2. Verify TDES elapses and the engine start signal is issued.
3. Observe the generator accelerate to rated voltage and frequency.
4. Verify TDNE elapses and the ATS transfers to emergency.
5. Confirm the load is served and the transition behaved as designed for the topology selected.
6. Restore the utility supply.
7. Verify TDEN elapses and the ATS retransfers to normal.
8. Verify TDEC elapses and the engine shuts down after its cooldown run.

NFPA 110 governs emergency and standby power system performance, and it is the standard behind the required testing intervals and the load acceptance requirements for life-safety systems. Facilities with life-safety loads have recurring test obligations that go well beyond commissioning.

Exam trap: A question describes an ATS serving large induction motors and asks why a delayed neutral transition is specified. It is not to protect the generator or to reduce inrush at the source — it is to let motor residual back-EMF decay so the motors are not reconnected out of phase with the oncoming source.