Bus transfer schemes — in-phase and residual-voltage transfer in MV switchgear

The guide on the ATS covers automatic transfer to a standby generator at low-voltage level. At medium-voltage level — for example between two independent MV supplies feeding an industrial installation or a building with a dual grid connection — a related but separate concept exists: the automatic bus transfer scheme, which automatically switches the load from one MV source to the other on loss of the active source.

The basic problem: two asynchronous sources

In a bus transfer scheme, the two MV sources (for example two independent supply transformers, or two grid connection points) are usually not mutually synchronised — unlike a permanently paralleled network, where synchronism is continuously monitored as covered in the guide on the synchronism-check relay (25). On loss of the active source, the breaker to that source must open and the breaker to the standby source must close — but the motor load on the bus (particularly large induction motors) continues to "coast down" for some time after the first breaker opens, generating a decaying residual voltage on the bus whose phase gradually drifts away from that of the new, incoming source.

If the second breaker closes at an arbitrary moment during that decaying residual voltage, an effectively unsynchronised paralleling occurs between the remaining motor EMF on the bus and the new source — comparable to the fault scenario that a synchronism-check relay normally prevents. This can produce a substantial current surge and mechanical shock loading on the connected motors and on the switchgear itself.

In-phase transfer (fast transfer)

In in-phase transfer, the protection continuously monitors the phase difference between the decaying residual voltage on the bus and the incoming new source, and closes the second breaker exactly at the moment both voltages are in phase (or within a very small, permissible phase band) — usually within tens to just over a hundred milliseconds after loss of the active source, i.e. before the residual voltage on the bus has decayed too far or drifted too far out of phase. Because the transfer takes place while the bus still has a substantial residual voltage that is moreover nearly in phase with the new source, the current surge remains limited and the connected motors keep running virtually without interruption.

In-phase transfer requires a sufficiently fast-closing breaker mechanism and a protection relay that can calculate the phase difference in real time and issue the close command with the correct lead time, accounting for the breaker's own closing time.

Residual-voltage transfer (slow, safe transfer)

If in-phase transfer is not achievable within the available time window (for example due to a slow breaker, or because the phase difference drifts too quickly), the scheme falls back to residual-voltage transfer: the second breaker closes only after the residual voltage on the bus has decayed below a set, safe threshold (typically on the order of 25-40% of rated voltage). This method thus deliberately waits until the motor EMF on the bus has decayed sufficiently that an unsynchronised closing no longer causes a damaging current surge — at the cost of a longer interruption for the load (typically several hundred milliseconds to a few seconds, depending on the coast-down behaviour of the connected motors).

Why both methods are often combined in one scheme

A complete bus transfer scheme usually attempts an in-phase transfer first, and automatically falls back to residual-voltage transfer if the in-phase conditions are not met within the allowed time window — for example because the phase difference builds up too quickly on a bus with many large, slowly coasting-down motors. A third, even slower fallback mode (a fixed time delay, independent of measured residual voltage) is sometimes applied as a last-resort option, at the cost of the longest interruption time but the greatest assurance against a damaging, unsynchronised closing.

Practical significance

The choice and setting of a bus transfer scheme is a trade-off between continuity (the shorter the interruption, the less disruption to critical processes and motors) and the mechanical/electrical stress on the switchgear and connected motors in the event of an incompletely synchronised closing. In installations with a large motor load on the same bus (where the motors' coast-down behaviour keeps the residual voltage present longer, but also causes it to drift out of phase faster), the in-phase detection window and the residual-voltage threshold must be tuned specifically to the characteristics of that motor load, not left at a generic factory setting.

Typical mistakes

Confusing a bus transfer scheme with a simple, fixed time delay — without active phase or voltage monitoring, there is no guarantee that the closing is actually safe at the moment the delay expires.
Setting the residual-voltage threshold too high, so the scheme effectively never achieves in-phase transfer and always waits unnecessarily long for the residual-voltage fallback.
Not adapting the in-phase detection window to the actual motor load on the bus — a bus with many large motors with slow coast-down behaviour has a different phase evolution than a bus with predominantly static (non-rotating) load.
Not configuring a fallback mode for the case where neither the in-phase nor the residual-voltage conditions are met within a reasonable time, leaving the transfer able to wait indefinitely without the load ever being restored.