Fault-Finding on Industrial Switchgear: A Systematic Diagnostic Guide

A tripped breaker on a production line creates immediate pressure: operations wants power restored now, and the instinct is to reset and re-energise. That instinct is how a transient nuisance trip and a developing insulation failure get treated identically — and how a fault that should have been caught at the panel becomes a repeat outage, or worse, a second failure with fault energy released.

This article is written for plant engineers, electrical technicians, and facility managers responsible for diagnosing switchgear trips at South African industrial sites. It sets out a systematic method for finding the actual root cause of a fault before re-energising, rather than treating a reset as the fix.

---

Before You Touch Anything: Isolation and PPE

No diagnostic step in this article assumes energised work. Isolate, lock out, tag out, and verify de-energisation with a rated voltage detector before any physical inspection of a tripped circuit begins — this is a baseline OHS Act and SANS 10142-1 obligation, not a discretionary precaution.

Where the trip involved a fault current rather than an overload, treat the switchgear as having absorbed real energy until proven otherwise. Vacuum interrupters, moulded-case breakers, and busbar joints can be damaged internally by a fault they successfully cleared, with no visible external sign. Do not re-energise a breaker that has interrupted a genuine fault current without first confirming its condition — see "When to Stop and Call for Backup" below.

---

Start With the Evidence: What Actually Happened

Before opening a single panel door, gather the evidence that already exists. This step is frequently skipped under time pressure, and skipping it is the single biggest cause of fault-finding that goes in circles.

Questions to answer first:

• Which device tripped, and what did its target/flag indicate? A protection relay's fault record — overcurrent, earth fault, differential — tells you what the relay saw, not necessarily the true cause, but it narrows the search dramatically.
• What was the load and site condition at the moment of trip? A trip during motor start points toward starting current and protection settings; a trip during a storm points toward transient overvoltage or lightning-induced surges; a trip that coincides with another team's cable or ground work is not a coincidence until ruled out.
• Is this the first trip, or a repeat? A breaker that has tripped on the same feeder twice in a month has a developing fault, even if it resets and holds each time. Treat repeat trips as more urgent than a single first-time trip, not less — normalisation of a "nuisance" trip is how sites end up with a catastrophic failure that "came out of nowhere."
• What do the SCADA or event logs show in the seconds before the trip? Voltage sag, current asymmetry between phases, or a sequence of alarms on adjacent equipment can point directly at the origin.

Only once this evidence is gathered does it make sense to open the panel.

---

Reading the Protection Relay Fault Record

On any switchgear with microprocessor-based protection relays, the fault record is the most information-dense piece of evidence available and should be read before any physical inspection.

What to extract from the record:

• Which protection element operated — overcurrent (ANSI 51), earth fault (ANSI 51N), differential (ANSI 87), or under/overvoltage — identifies the fault type.
• Fault current magnitude and duration. A fault current close to the relay's pickup threshold that cleared slowly suggests a marginal or developing fault. A fault current at or near the calculated maximum fault level, cleared quickly, is more consistent with a solid short circuit.
• Phase involvement. Single-phase-to-earth faults, phase-to-phase faults, and three-phase faults each point toward different physical causes — cable insulation breakdown tends to start single-phase-to-earth, while mechanical damage to a cable or busbar often produces multi-phase faults.
• Pre-fault current trend, if the relay logs it. A load that was trending upward before the trip suggests overload rather than a sudden insulation failure.

Compare what the relay recorded against the protection coordination study for the installation. A relay that operated outside its expected time-current characteristic, or that should not have seen fault current for a fault located elsewhere on the network, indicates either a coordination problem or a fault location different from the obvious one.

---

The Physical Inspection Sequence

With the evidence reviewed and the switchgear isolated, locked out, and verified dead, work outward from the point indicated by the fault record.

1. Visual inspection first, before any test instrument is connected. Look for carbon tracking, discolouration, physical damage to insulation, evidence of moisture or vermin ingress, and any smell of burnt insulation. Photograph everything before disturbing it — this record matters for both root-cause analysis and any insurance or warranty process.
2. Insulation resistance (IR) test the suspect circuit to confirm whether the insulation has actually failed or the trip was a protection operation without genuine insulation breakdown (for example, a nuisance trip from a maloperating relay, a loose CT connection, or a genuine overload that never became a dielectric failure).
3. Contact resistance (Ductor/DLRO) test the breaker and adjacent joints if the fault record shows a fault current was interrupted. Elevated contact resistance after a fault-clearing operation indicates contact erosion from the interruption itself.
4. Check CT and VT wiring and burden when the relay's reading is inconsistent with the physical evidence — a genuinely healthy circuit that shows a phantom fault on the relay is frequently a wiring or CT saturation problem, not a network fault at all.
5. Torque-check terminations in the affected section. A large proportion of LV fault trips trace back to a loose termination that ran hot, tracked, and eventually flashed over — this is found by physically re-torquing and inspecting, not by electrical testing alone.

---

Common Fault Patterns and Where to Look

This table is a starting point for narrowing the search, not a substitute for testing — several of these symptoms overlap, and only measurement confirms the actual cause.

---

MV-Specific Diagnostic Considerations

Medium-voltage fault-finding carries additional risk and additional diagnostic steps beyond LV practice.

Vacuum interrupter integrity after a fault-clearing operation. A VCB that has just interrupted a genuine MV fault should not be assumed serviceable because it operated correctly. A high-potential dielectric test of the vacuum bottle, or at minimum a contact resistance and mechanical timing check, should be performed before the breaker is returned to service.

SF6 gas pressure on gas-insulated switchgear and ring main units. A fault interruption can, in some failure modes, correlate with gas pressure loss. Check gas pressure against the manufacturer's specification as part of the post-fault sequence, not only on the routine maintenance cycle.

Cable fault location before excavation or re-termination. For MV cable faults, a time-domain reflectometry (TDR) or surge-pulse cable fault locator should be used to pinpoint the fault location before any excavation or cable replacement is attempted — guessing the location from route drawings alone leads to unnecessary excavation and extended downtime.

---

Tools of the Trade

A well-equipped fault-finding response does not require every instrument on every call-out, but the following cover the large majority of industrial switchgear faults:

• Rated voltage detector for safe isolation verification
• Insulation resistance tester (Megger) — 1000 V DC for LV, 2500 V DC or higher for MV
• Micro-ohmmeter (Ductor/DLRO) for contact and joint resistance
• Infrared thermal camera for locating hotspots under load, once the circuit is restored to service
• Relay test set for confirming protection operation against the coordination study
• Clamp meter with phase-angle capability for load and current-balance checks
• Cable fault locator (TDR/surge) for MV cable fault pinpointing

---

When to Stop and Call for Backup

Not every fault should be diagnosed and cleared by the on-site team alone. Escalate to specialist support (an OEM service partner, a protection engineer, or GNTech's technical team) when:

• A repeat fault on the same circuit has not been resolved by the initial corrective action
• An MV breaker has interrupted a fault current and its post-fault condition is uncertain
• The fault record is inconsistent with the physical evidence and the discrepancy cannot be resolved on site
• The protection coordination study is out of date relative to the current network configuration
• Fault energy or incident-energy exposure at the point of investigation exceeds what the site's PPE rating and procedures cover

Calling for backup at this point is not a failure of the site team — it is the correct application of a fault-finding process that recognises its own limits.

---

Closing the Loop: Documentation and Repeat-Fault Prevention

A fault investigation that ends with "reset and monitor" without a documented root cause is not closed — it is deferred. Every fault investigation should produce a written record covering: the fault record data, the physical findings, the test results, the corrective action taken, and — critically — whether the protection coordination study or maintenance programme needs to be updated as a result.

Faults that recur on the same feeder within a short period should trigger a review of the maintenance schedule and coordination study for that circuit, not just a repeat of the same reactive fix.

---

Working With GNTech

GNTech supplies arc-resistant switchgear, Pronutec MV ring main units, Telergon LV isolation and protection equipment, and spare parts and contact sets for fault repair and replacement. Our technical team can assist with fault investigation support, protection coordination studies, and post-fault equipment assessment for South African industrial sites. Contact the GNTech team to discuss fault-finding support or a maintenance programme that reduces repeat failures.