10.3 Fuses, GFCI & Zero-Sequence CT Grounding

Fuse Continuity and Resistance Test

A fuse is the simplest protective device - a sacrificial element that melts when current exceeds its rating. But a fuse can fail open (blown element), fail partially (hairline crack giving high resistance), or in the case of current-limiting fuses, have an element that has shifted characteristic after a near-fault. ANSI/NETA ATS requires a continuity or resistance check on every fuse at acceptance.

Procedure

1. De-energize the circuit and verify dead with a voltage tester. Apply lockout/tagout.
2. Remove the fuse from its clips using insulated fuse pullers. Testing in place gives false readings because parallel fuses, buswork, or transformer windings provide alternate current paths around the fuse under test.
3. Set a DMM to ohms or continuity and measure across the fuse ends. A good power fuse reads near 0 ohm (typically a few milliohms for large current-limiting fuses; small signal fuses read a fraction of an ohm). A reading above the manufacturer's published value, or an open, means replace the fuse.
4. Visually inspect the fuse body for cracks, discoloration, sand leakage (in current-limiting fuses), and ejector/signal indicators (SMD-style fuses have a pin that pops when blown).
5. For high-voltage or current-limiting fuses, also verify the catalog number and ampere rating match the coordination study - a common field error is installing an RK5 in a location that the one-line specifies as RK1.

Fuse Class Quick Reference

RK1 provides lower peak let-through energy than RK5, giving better arc-flash and component protection when the coordination study allows it.

GFCI (Class A) - Trip Levels and Functional Testing

A Ground Fault Circuit Interrupter (GFCI) protects personnel by comparing current going out to the load with current coming back. Any difference is leakage to ground - current taking an unintended path through a person, a tool chassis, or wet conductors. The sensing element is a core-balance current transformer (CBCT) that all current-carrying conductors pass through.

UL 943 Class A Trip Threshold

Per UL 943 (Edition 5, latest revision August 25, 2025), a Class A GFCI must:

• Trip at 6 mA or more of ground-fault current.
• NOT trip at 4 mA or less (prevents nuisance trips from normal appliance leakage).

So the operating window is 4-6 mA, often stated as a nominal 5 mA threshold. Higher settings exist for other applications: 30 mA for European RCDs, 300 mA or higher for equipment protection (GFPE) - these are not Class A personnel protection devices.

Trip Time

UL 943 defines a trip-time curve. Key points:

The 25 ms point at 264 mA is the number most often cited - but it is not the trip time at the 5-6 mA threshold, where UL allows several seconds. Do not confuse the two on the exam.

Functional Testing Procedure

ANSI/NETA requires GFCI functional testing at acceptance (pre-energization) and periodically in service:

1. Built-in Test button - pressing it simulates a small ground fault (~6 mA) inside the device. The GFCI must trip. This verifies the sensing toroid, electronics, and trip solenoid as a complete chain.
2. Calibrated external GFCI tester - a plug-in tester that injects a precise ground-fault current (commonly 5, 8, or 30 mA) between hot and equipment ground, measuring both trip current and trip time. Use this for acceptance documentation.
3. Since June 29, 2015, UL 943 also requires an auto-monitoring (self-test) function inside the device that periodically checks the electronics. The manual Test button is still required because the self-test cannot verify the trip solenoid or welded contacts - only an actual trip does that.

A GFCI that does not trip when the Test button is pressed is failed and must be replaced. A device that trips but at the wrong current or too slowly (per the calibrated tester) is also failed.

Zero-Sequence (Core-Balance) CT for Ground-Fault Sensing

In industrial and medium-voltage switchgear, ground-fault protection (ANSI device 50G/51G) often uses a zero-sequence CT - also called a core-balance or window CT - instead of residual sensing from three phase CTs.

How It Works

All phase conductors (and the neutral, if present) of the protected circuit pass through a single window-type CT. In a healthy circuit, the phasor sum of the currents is zero - the magnetic fields from the conductors cancel inside the core, and the CT secondary produces no output. When a ground fault occurs, some current leaves the circuit via the fault path (equipment grounding conductor, conduit, earth) and does not return through the CT window. The sum is no longer zero; the CT secondary delivers a current proportional to the residual (ground-fault) current, and the relay trips.

Placement and the Neutral Problem

The most common wiring error with a zero-sequence CT is mishandling the neutral and shields:

• Neutral must pass through the window on a 4-wire system, in the same direction as the phases. If it is routed outside the CT, normal unbalanced load current on the neutral looks like ground-fault current and causes false trips.
• Cable shields and grounding conductors must either bypass the CT window entirely or pass through the window and back in the opposite direction. If a shield simply passes through along with the phases, shield current cancels part of the true ground-fault current and the relay under-trips.
• All conductors of the same circuit must pass through the same window - do not mix circuits in one CT.

Zero-Sequence vs. Residual (50/51G) Sensing

Zero-sequence sensing is preferred for low-level ground-fault detection (e.g., 50G instantaneous) because it avoids the mismatch errors inherent in summing three separate phase CTs.