Capacitor-bank unbalance protection (ANSI 60C/59N) — fuseless, internally and externally fused

The guide on capacitor-bank discharge resistors covers how a capacitor bank is safely discharged after switching off, and the guide on detuned reactors covers resonance risk with harmonics. This article covers a different question: how is the failure of an individual capacitor element or unit within an operating bank detected, long before it leads to a complete bank failure? The answer is unbalance protection, commonly designated with the ANSI codes 60C (unbalance current) or 59N (neutral-point/unbalance voltage), and worked out following the methodology of IEEE C37.99 ("IEEE Guide for the Protection of Shunt Capacitor Banks").

Why ordinary overcurrent protection falls short

A capacitor bank draws a virtually constant, stable current at rated voltage. The failure of a single element or a single unit within a large bank changes the bank's total current only marginally — too small to be picked up by conventional overcurrent protection (ANSI 51). Yet that failure is still relevant: the remaining, healthy elements in the same series group must then withstand a higher voltage per element, increasing the likelihood of a subsequent, faster cascade of element failures. Unbalance protection therefore does not detect the bank's absolute current or voltage, but the difference (the unbalance) between parts of the bank that would be identical in a healthy bank.

Three bank configurations, three detection principles

Externally fused: each capacitor unit has its own fuse mounted outside the unit. When a unit fails, its fuse blows and removes the entire unit from service. The resulting unbalance — for example a shift in neutral-point voltage on an ungrounded, wye-connected bank — is measured with a voltage relay between the neutral point and ground (often designated 59N), or with a current transformer between two bank halves.
Internally fused: each individual element within a unit has its own internal fuse. When an element fails, only that internal fuse blows; the unit itself stays in service with a slightly higher voltage across the remaining, parallel-connected elements in that group. Because there is no external fuse operation giving a visible signal, the cumulative effect of multiple internal element failures is tracked via an unbalance-current relay (often called 60C) in a bridge circuit between two parallel bank sections.
Fuseless: here there are no fuses at all — each element is designed to fail into a low-impedance short (an internal weld) rather than open-circuiting, so the series string stays in service with the remaining elements, albeit with a slightly altered net capacitance and voltage distribution. Because there is no fuse operation that can be signalled separately, unbalance protection on a fuseless bank is the only way to detect a failed element.

Alarm and trip thresholds

Unbalance protection typically works with two levels:

An alarm level, set at an unbalance corresponding to the failure of the first few elements or units — at this level the bank still remains well within the voltage and thermal limits of the remaining, healthy elements, but the alarm signals that maintenance or replacement should be scheduled.
A trip level, set at an unbalance indicating that further element failures would push the voltage across the remaining elements close to their dielectric or thermal limit — at this level the bank trips to prevent a cascade of rapidly successive element failures (and ultimately an internal short-circuit failure of the bank).

Note: the exact unbalance threshold must account for the inherent manufacturing tolerance between capacitor elements — a new, fully healthy bank already shows a small, measurable unbalance due to normal production spread. A too-sensitive unbalance setting leads to nuisance alarms or trips; a too-insensitive setting misses early degradation. IEEE C37.99 provides calculation methods to correctly dimension this sensitivity per bank configuration.

Practical relevance

When assessing an existing capacitor bank, it is important to establish which of the three configurations (externally fused, internally fused, fuseless) has been applied, because this directly determines which detection principle (physical fuse indication versus unbalance measurement alone) is available — and therefore how an operator can recognise a degrading bank before it fails completely or develops an internal fault.

Common mistakes

Relying only on ordinary overcurrent protection (51) to detect element failure — it only responds to a gross bank fault, not to the failure of individual elements or units.
Not accounting for the bank's manufacturing tolerance when setting the unbalance threshold, leading to nuisance alarms or, worse, an over-insensitive setting that misses early degradation.
Confusing unbalance detection on a fuseless bank with fuse-blown indication on a fused bank — a fuseless bank has no separate fuse signal; unbalance protection is the only detection method there.
Ignoring temperature-driven capacitance drift when establishing the unbalance baseline — capacitor element capacitance varies somewhat with temperature, which can cause a small, harmless unbalance that should not be mistaken for element failure.