In an arc-flash matter the physics is often the least contested part. Incident energy can be calculated, the burn pattern examined, the duration bounded. What decides responsibility is usually a paper trail: whether an analysis existed, whether it was current, whether the label reflected it, and whether the work being performed was authorised to be performed energized at all.
What the standard actually requires
NFPA 70E requires an employer to determine the arc-flash hazard and document it, either through an incident energy analysis or the arc-flash PPE category method, and to field-mark equipment likely to require examination or servicing while energized. The label must convey nominal voltage, the arc-flash boundary, and either incident energy at a stated working distance or the PPE category.
The standard also requires that energized work be justified rather than merely convenient, and where it is justified, authorised through an energized electrical work permit. Those two documents — the label and the permit — are the ones most often missing.
Studies go stale quietly
An analysis is a snapshot of an electrical system at one moment. Available fault current changes when a utility upgrades a transformer, when a service is enlarged, when generation is added, or when a plant reconfigures. Incident energy moves with it, and not always in the direction intuition suggests — reducing fault current can lengthen clearing time and raise incident energy.
So the useful question is not whether a study exists but when it was performed and what has changed since. A five-year review interval is a common practice reference point; a study predating a documented service change is a finding on its own.
Labels people can actually use
A technically correct label that a worker cannot act on has not done its job. Boundaries that assume a working distance different from the actual task position, labels applied to an enclosure but not to the specific compartment being worked, and labels stating a PPE category without the corresponding equipment available on site are all recurring gaps.
The physical state of the label matters too. Sun-faded, painted over, or removed during a refurbishment are ordinary occurrences and are worth photographing before anything is cleaned up.
Justification for energized work
The standard's default is de-energized work; energized work requires that de-energizing would introduce additional hazards or be infeasible due to equipment design or operational limitations. Schedule pressure and production loss are, deliberately, not on that list.
Where energized work was performed, the permit and its stated justification are the record of the decision. Where no permit exists, the question becomes who decided, on what basis, and whether the decision was within their authority — which shifts the matter from an electrical question to a management one.
PPE, and what it is rated to do
Arc-rated clothing is specified by arc thermal performance value, and the rating describes protection against thermal energy at a stated incident energy. It is not protection against the pressure wave, against molten metal ejection, or against an event exceeding the rating.
Examination of the garments actually worn is informative: layer construction, whether non-arc-rated synthetics were worn underneath, the condition and laundering history, and whether the ensemble was complete. Burn distribution frequently maps onto gaps in the ensemble rather than failures of the fabric.
Training and qualification
The standard distinguishes qualified from unqualified persons in terms of demonstrated skill and knowledge of the specific equipment and its hazards, not job title or years served. Training records, the content of that training and its retraining interval are documentary questions with documentary answers.
A worker directed to perform a task they were not qualified to perform is a different matter from a qualified worker encountering an unforeseeable condition, and the records usually distinguish the two clearly.
Reconstruction with the documents
The reconstruction that holds up assembles both halves: the physical evidence — equipment condition, burn pattern, device state, arc duration — and the documentary record of what was known, labelled, authorised and trained. Where the two agree, the account is durable. Where the physical severity exceeds what the documents anticipated, the gap is the finding.
This is also why early document preservation requests matter as much as equipment preservation. Studies, labels, permits, training records, maintenance history and relay data all have their own retention practices, and several are routinely superseded rather than archived.
Who owns the obligation
Arc-flash obligations attach to the employer of the person doing the work, which becomes genuinely complicated on a site with a host employer, a contractor and possibly a sub-tier. NFPA 70E addresses that directly: the host must inform the contractor of known hazards and of the documented incident energy, and the contractor must instruct its own people.
That exchange is supposed to leave a record, and the absence of one is a common finding on contractor injuries. Where a contractor was never given the study or the labels were absent, responsibility moves upstream regardless of who was holding the tool.
The recurring conclusion
Arc-flash incidents are rarely caused by an absence of knowledge about arc flash. The mechanisms are well characterised and the standards are mature. They are caused by a study nobody refreshed, a label nobody updated, a permit nobody wrote, or a decision to work energized that nobody documented.
That is an uncomfortable finding for everyone involved, and it is also the reason these matters are usually decided on paper.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.