Power-swing blocking (ANSI 68) — stopping distance protection from misreading a stable swing as a fault
Power-swing blocking (ANSI 68) — stopping distance protection from misreading a stable swing as a fault
The pole-slip / out-of-step protection guide (ANSI 78) covers the relay function that actually trips a generator once it has genuinely lost synchronism. This article covers its close counterpart, ANSI device number 68: a function that does the opposite of tripping — it blocks other, faster protection functions (most often distance protection zones) from tripping unnecessarily while the system passes through a stable power swing that is not actually a fault.
Why a stable swing looks like a fault to a fast relay
During and after a disturbance elsewhere in the network (a fault that has already been cleared, a large load or generation step, a switching operation), the power angle between two parts of an interconnected system can oscillate before settling back to a new equilibrium — a power swing. While that oscillation is under way, the impedance measured by a distance relay at a given location moves through the R-X plane in a way that can pass directly through, or very close to, the relay's own tripping characteristic — exactly as a genuine internal fault would. Without a dedicated function to tell the two apart, a distance relay's faster zones can trip on a swing that was never actually a fault, needlessly disconnecting a healthy line and potentially worsening the very disturbance the system was recovering from.
Telling a swing from a fault: the rate of change, not the position
The key distinguishing feature is not where the impedance ends up, but how fast it gets there. A genuine fault causes an essentially step-change in impedance, reaching the relay's characteristic within one cycle or so. A power swing, by contrast, evolves over a comparatively long timescale — typically many tens to hundreds of milliseconds — as the power angle between the two systems rotates. Power-swing blocking exploits exactly this difference in speed.
How the blocking function works: dual blinders and a timer
A power-swing blocking function typically uses two concentric impedance characteristics (often called the outer and inner "blinders", sometimes drawn as parallel straight lines either side of the protected line's characteristic, sometimes as concentric circles or polygons):
- As the measured impedance moves inward, it first crosses the outer blinder, which starts a timer.
- If the impedance then reaches the inner blinder before a preset time threshold Δt has elapsed, the transit is judged too fast to be a swing — this is treated as a genuine fault, and the normal distance-protection zones are not blocked.
- If the transit from outer to inner blinder takes longer than Δt, the movement is judged too slow to be a fault — this is classified as a power swing, and the selected distance zones are blocked from tripping for as long as the impedance remains inside the swing detection region.
Note: the exact timer threshold Δt and blinder spacing are calculated per installation, based on the fastest credible swing rate the local grid can produce and the slowest credible internal fault that must still be cleared without delay — there is no single universal value that applies to every installation.
Why zone 1 is often deliberately left unblocked
Practice commonly excludes the instantaneous first distance zone (see the distance-protection zones guide) from power-swing blocking, and blocks only the slower, time-delayed zones (typically zone 2 and zone 3). The reasoning: zone 1 is already set with a conservative reach that should, in normal operation, only ever see a genuine close-in fault, not a swing — so leaving it unblocked preserves the fastest possible clearance for a real fault while the blocking logic protects the more far-reaching, and therefore more swing-exposed, back-up zones.
The unblocking problem: a fault during an already-blocked swing
Blocking cannot simply remain in force indefinitely once triggered: a genuine fault can develop while a swing is already in progress and already blocked. A robust power-swing blocking scheme therefore includes an unblocking (or "out-of-step trip permit") path that recognises this evolving-fault condition — typically by monitoring for the sudden appearance of negative- or zero-sequence quantities (a symmetrical three-phase swing produces essentially none, while almost any unbalanced fault does) or an abrupt, fault-like discontinuity in the measured quantities during the swing — and permits tripping despite the block still nominally being active. Many schemes also apply a maximum blocking duration as a further safeguard, so that blocking cannot persist indefinitely if the swing itself fails to resolve.
Practical relevance
When reviewing the protection settings of an interconnected line or a large embedded generator (see also the related pole-slip protection guide), it should be verified that power-swing blocking is present on the appropriate distance zones, that the blinder spacing and timer are actually derived from a stability study of the specific network rather than copied from a default setting, and that an unblocking path for a fault-during-swing condition is implemented — a blocking function without an unblocking path can delay clearance of a genuine fault that happens to coincide with a swing.
Common mistakes
- Confusing power-swing blocking (68) with pole-slip protection (78) — the former prevents an unnecessary trip during a stable condition, the latter deliberately trips once the system has genuinely lost synchronism; see the related guide for that distinction in more detail.
- Setting the blocking timer threshold too long or too short — too short a threshold can also block on a genuinely fast internal fault; too long a threshold can fail to block a fast swing.
- Blocking zone 1 along with the back-up zones, unnecessarily delaying clearance of the close-in faults that zone 1 exists to clear instantaneously.
- Omitting the unblocking path for a fault developing during an already-blocked swing, leaving a genuine fault under that condition to be cleared only by slower, back-up protection instead of the intended zone.
Related
Further reading
- ANSI 62 (pole discrepancy)Pole discrepancy protection (ANSI 62PD) — when not all poles of a circuit breaker switch together
- ANSI 85 (teleprotectie)Teleprotection schemes (ANSI 85) — permissive and blocking with distance protection
- Praktijk (ANSI 21)Distance protection (ANSI 21) — impedance protection with zone 1/2/3 on MV and HV lines
- Praktijk (ANSI 67, richtingsrelais)Directional overcurrent protection (ANSI 67) — why an ordinary overcurrent relay falls short on a ring network or double-fed busbar
- Praktijk (ANSI 32, generator)Reverse power protection (ANSI 32) — why a generator that keeps turning is no proof that all is well
- Praktijk (ANSI 86)Lockout relay (ANSI 86) — why a protection trip does not clear itself but requires a manual reset