An arc flash lasts milliseconds and can end a career, a life, and a facility. Determining what initiated it — and why protection did not clear it in time — is forensic work.
Start a conversation with our AI Research Concierge, already scoped to arc flash & arc fault. Pick a starting point, or describe your situation directly.
An arc flash is a short, violent event: temperatures at the arc terminals can approach 19,000 °C, pressure waves are measured in kilopascals, and the whole thing is usually over inside a few cycles. The damage that remains is severe but not random. Copper vaporized by an arc deposits differently than copper melted by a fire; enclosure deformation records the pressure rise; and the protective device that should have cleared the fault carries its own record of whether it operated, operated late, or never operated at all. The forensic question is almost never "was there an arc" — it is what initiated it, how much incident energy was released, and whether the protection scheme and work practices met the standard of care.
An arc needs an initiating event that bridges or breaks down the insulation between conductors. Identifying which one applies is the core of the investigation.
A dropped tool, probe slip, or misrouted conductor creating a phase-to-phase or phase-to-ground path during energized work.
Dust, moisture, or salt films creating carbonized tracking paths that progressively reduce dielectric strength until breakdown.
High-resistance joints generating heat, oxidizing further, and eventually breaking down into a series arc fault.
Thermal aging, mechanical damage, rodent or vibration damage, or contamination degrading insulation to failure.
Breakers, relays, or fuses mis-set, mis-selected, or failed — extending arc duration and multiplying incident energy.
Racking mechanism faults, worn contacts, missing barriers, or deferred maintenance creating the conditions for a fault.
Arc flash reconstruction combines physical evidence from the equipment with electrical modelling of what the system could deliver.
A single arc flash event routinely puts several of these in motion at once:
The switchgear, conductors, PPE, and — critically — the breakers and fuses are the evidence. Re-energizing, cleaning, or scrapping them can destroy the proof of what failed and when.
An arc flash begins when insulation or air between conductors breaks down and current flows through ionized air. Common initiators are a tool or conductor bridging energized parts, contamination and surface tracking, a loose high-resistance connection degrading into a series arc, aged or damaged insulation, and equipment defects such as worn contacts or faulty racking mechanisms. Which one applies is determined from the physical evidence, not the description.
Incident energy is modelled under IEEE 1584 using the available bolted fault current, the arc duration set by how quickly protection cleared, the working distance, the enclosure geometry, and the system voltage. In a forensic setting the modelled result is then tested against reality — the burn pattern, PPE damage, and equipment damage should be consistent with the calculated energy. When they are not, that discrepancy usually points to the real story, often an arc duration far longer than design assumptions.
Usually yes. Breakers, trip units, relays, and fuses can be tested to verify whether they operated within their published curves, and protective-relay and power-quality records often timestamp the fault and the clearing. A device that was mis-set, mis-coordinated, previously damaged, or simply failed to operate extends arc duration — and incident energy scales roughly with duration, so a clearing failure often matters more to the outcome than the initiating event.
Preserve the entire circuit path and its protection: the equipment and enclosure as found, conductors, the breakers, fuses and relays, plus the damaged PPE and clothing. Also secure electronic records — relay events, power-quality data, SCADA and maintenance history — which are frequently overwritten within days. Do not re-energize, clean, repair, or scrap anything before it is documented.
NFPA 70E governs electrical safety in the workplace including risk assessment, labeling, and PPE selection. IEEE 1584 is the accepted method for calculating incident energy. NESC applies to utility installations, and OSHA 1910 Subpart S sets the regulatory baseline. Investigations typically evaluate both the physical failure and conformance with these requirements.
Technical briefings from our work in this area.
Most arc-flash matters turn less on physics than on documents: whether the study was current, the label correct, the boundary right, and the energized work authorised.
readIncident energy scales with how long the arc burns, so the protective device is usually the difference between an incident and a catastrophe. Its as-found condition is evidence.
readIncident energy is calculated under IEEE 1584 from assumptions about fault current, arc duration, and working distance. When the burn pattern does not match the number, the discrepancy is usually the most important evidence in the case.
readTell us what occurred. We will triage it and connect you with the right expert — usually within one business day.