Can arcing damage show where a fire started?

It can help. Arc mapping records where arcing occurred on each circuit; because an energized circuit usually arcs first where the fire first reaches it, the pattern across several circuits can point back toward the area of origin. It is one line of evidence among several, and it has to be consistent with the fire patterns, the witness accounts, and the recorded data.

What if the cause cannot be determined?

Then the report should say so. Under NFPA 921, when two or more hypotheses cannot be ruled out with the available evidence, the cause is undetermined. An opinion that names a probable cause the evidence does not support is exactly what a reliability challenge is designed to exclude.

Does PEFG replace the fire investigator?

No. On most fire matters PEFG works alongside the origin-and-cause investigator, answering the electrical questions: whether a circuit or device was capable of ignition, whether the arcing is cause or effect, and what the protective devices and recorded data show.

Electrical Fire & Arc Flash Incident Expert

Electrical activity: cause or casualty?

Fire damages wiring. Arcing found at a fire scene may be what started the fire, or it may be what the fire did to energized conductors on its way through. Telling the two apart, through arc mapping, the sequence in which circuits failed, and the state of every overcurrent device, is the central electrical question in most fire matters. PEFG typically works alongside the origin-and-cause investigator, answering the electrical questions the fire investigation raises.

The method

NFPA 921 calls for the scientific method, and courts frequently look to NFPA 921 when a fire opinion is challenged:

Recognize the need, and define the problem.
Collect data: physical, documentary, recorded, and witness.
Analyze the data.
Develop hypotheses (inductive reasoning).
Test each hypothesis against all of the data (deductive reasoning).
Select the final hypothesis, only one that survives testing.

Opinions are stated to a level of certainty. A probable cause is more likely than not; a possible cause is not. When two or more hypotheses cannot be eliminated with the evidence available, undetermined is a valid conclusion, and a more defensible one than a forced choice.

Three errors a defensible investigation avoids

Presumption: arriving with a cause in mind before the evidence has produced testable hypotheses.
Expectation bias: reaching a conclusion before all of the relevant data has been examined.
Confirmation bias: relying on the data that supports a hypothesis while ignoring or dismissing the data that does not.

The investigation file should show, step by step, that each was avoided.

What gets examined

Arc beads, melting, and conductor damage, mapped across every circuit in the area of origin.
The state of every breaker and fuse, and whether each operated.
Terminations, connections, receptacles, and equipment internals near the origin.
Recorded data: relay and meter events, alarm logs, and video where it exists.
Laboratory work where warranted: optical and electron microscopy, energy-dispersive X-ray spectroscopy, CT and radiography first, and cross-sectioning or other destructive testing only under a protocol agreed by all parties.

Arc flash incidents and injury claims

When an arc flash injures a worker, two questions usually run in parallel: why the equipment failed or the arc was initiated, and whether the incident energy, arc flash boundary, labeling, and PPE were what IEEE 1584 and NFPA 70E required. The second question needs the system configuration as it was at the moment of the incident, not as it appears in the current study. See arc flash analysis.

A published case study: arc flash inside air-insulated switchgear

From a case study Jay Prigmore authored for a co-presented 2024 IEEE IAS Electrical Safety Workshop tutorial. A teaching example, not a client matter.

After six years in service, a medium-voltage air-insulated switchgear lineup failed at the end of an A-phase voltage-transformer bushing, and video captured the arc flash. The recording shows the fault progressing from A-phase-to-ground, to A-B phase-to-phase, to a three-phase fault that continued until cleared.

The bushing was cast cycloaliphatic epoxy with an inner high-voltage screen and an outer grounded screen. On the failed part the two screens were not concentric: the smallest spacing between them sat near the failure location, and a void was observed in the casting.

The manufacturer’s stated dielectric strength of 330 V per mil sets the margin. The roughly 15.6 kV phase-to-ground stress of a 27 kV solidly grounded system needs only about 47 mil of epoxy, but withstanding the 125 kV impulse level of that voltage class takes about 379 mil, close to 3/8 in., which is the design spacing. A rough measurement of the thinnest section came out just under 10 mm, at the minimum for the impulse level, before accounting for the off-center screens and the void.

The root cause was a poorly manufactured bushing with voids in the insulating material. A subsequent audit at the manufacturer found that production staff had not followed the manufacturing instructions and had deviated from the specified steps. Those deviations, together with poor control of the vacuum process meant to extract air pockets and of the concentric spacers that position the internal screens, produced the voids and the off-center screens that led to the failure. As of the 2024 presentation, corrective actions were in place and no further failures had been reported.

Methodology and standards applied

NFPA 921, Guide for Fire and Explosion Investigations: the scientific method, arc mapping, and the elimination of alternative hypotheses.
ASTM E1188 (collection and preservation of evidence), ASTM E860 (examining items that are or may become involved in litigation), and ASTM E620 (reporting expert opinions).
IEEE 1584 and NFPA 70E: incident energy, arc flash boundary, and PPE.
NFPA 70 (NEC): in the edition adopted when the installation was made.

Before anyone disturbs the scene

Put all potentially interested parties on notice, so each can investigate and none can claim the examination was unfair.
Hold a joint scene inspection, largely non-destructive, before any evidence is removed.
Do not clean, re-energize, or disassemble; photograph everything as found, with a scale.
Remove evidence under chain of custody, and leave destructive testing for an agreed laboratory protocol.

Download the printable checklist and evidence log (PDF, 2 pages) →

Electrical Fire & Arc Flash Incident Investigation