Understanding Arc Flash Hazard Analysis for Commercial Projects

Arc flash is among the most severe hazards in commercial electrical work and among the most poorly understood by the people who own the equipment. An arcing fault releases enormous energy in milliseconds — intense heat, blinding light, a pressure wave, and vaporized metal — and the injuries are catastrophic and frequently fatal.

An arc flash hazard analysis quantifies that risk at each piece of equipment so it can be managed with engineering and procedure rather than guesswork. This is an overview of what the study involves and why it matters well beyond the day it's completed.

What an arc flash is

An arc flash occurs when current jumps through air between conductors or to ground. The air itself becomes a conductor — an ionized plasma reaching temperatures several times hotter than the surface of the sun.

The damage comes from several mechanisms at once. Radiant heat causes severe burns at distances that surprise people. The pressure wave — the arc blast — can throw a worker across a room and cause concussive and hearing injury. Molten metal is ejected at high velocity. The light is intense enough to cause lasting vision damage.

Common initiating causes are mundane: a dropped tool bridging conductors, insulation degraded by age or rodents, moisture or dust accumulation, a loose connection heating over time, or equipment failing during switching.

Incident energy and why time dominates

The central quantity is incident energy — thermal energy delivered to a surface at a given distance, expressed in calories per square centimeter. It determines what protection is required.

Incident energy depends on available fault current, the distance from the arc, and — most importantly — how long the arc lasts before a protective device clears it.

That last variable dominates, and it produces a result many people find counterintuitive: lower available fault current can produce higher incident energy. A lower current may not trip a breaker into its instantaneous region, so the arc burns for a full second instead of a few cycles. Duration matters more than magnitude, which is why intuition is a poor guide and why the study is necessary.

What the study involves

A proper analysis is a sequence, and each stage depends on the one before.

Data collection

The foundation, and the most labor-intensive part. It requires utility fault current and impedance data, transformer ratings and impedances, conductor sizes and lengths, and the make, model, and settings of every protective device. Field verification is essential — existing drawings are frequently wrong, and a study built on inaccurate data produces confidently incorrect labels.

Short circuit study

Available fault current is calculated at each point in the system. This also verifies that equipment interrupting ratings are adequate, which occasionally surfaces problems that have nothing to do with arc flash.

Protective device coordination

The study establishes how devices are set to operate — ideally so the nearest upstream device clears a fault while everything else stays closed. Coordination and arc flash mitigation frequently pull in opposite directions: faster tripping reduces incident energy but can sacrifice selectivity. Resolving that tension is engineering judgment.

Incident energy calculation

Using IEEE 1584 methodology, incident energy and approach boundaries are calculated at each location based on the results above.

Labeling and documentation

Equipment is labeled per NFPA 70E, and the study is delivered with single-line diagrams, calculations, device settings, and recommendations.

Reading the label

A compliant label typically carries nominal system voltage, incident energy at working distance, the arc flash boundary, required PPE, and the shock approach boundaries with glove class.

Two points about interpretation. The arc flash boundary is the distance at which incident energy falls to 1.2 cal/cm² — the threshold for a second-degree burn on bare skin. It is not a safe line; it is the distance at which an unprotected person is merely injured rather than severely so.

And the incident energy figure applies at a specific working distance, usually 18 inches. Energy rises sharply as distance decreases. Reaching further into equipment than the assumed distance invalidates the PPE selection the label supports.

Reducing the hazard by design

The study identifies risk; design reduces it. Options, roughly in order of effectiveness:

• De-energize. The only complete protection. NFPA 70E treats energized work as requiring justification, not preference.
• Remote operation. Remote racking and remote switching move the worker outside the boundary entirely.
• Arc energy reduction. Maintenance switches that temporarily lower trip settings during work, zone selective interlocking, differential protection, and optical arc detection that clears in a few milliseconds.
• Current limiting devices. Fuses and breakers that limit let-through energy.
• Arc-resistant equipment. Switchgear designed to vent arc energy away from personnel.
• PPE. Last, not first. It mitigates injury; it does not prevent the event.

The study is not permanent

This is the part most often missed. An arc flash study reflects the system as it existed when the data was gathered. It goes stale.

NFPA 70E calls for review at intervals not exceeding five years, and sooner whenever the system changes materially — a service upgrade, a new transformer, added distribution, replaced protective devices, or a change in utility fault current from grid-side work you may not be notified about.

There is also a maintenance dependency that deserves more attention than it gets. Calculations assume protective devices operate as specified. A breaker that has not been exercised or tested in twenty years may not clear in the time the study assumed — and every label downstream is then optimistic. Protective device maintenance is part of arc flash safety, not a separate program.

What owners should ask for

If you're commissioning a study, the deliverable should include field-verified data rather than drawing-derived assumptions, complete short circuit and coordination analysis, incident energy at every relevant location, labels installed on equipment, an editable model you retain, and specific mitigation recommendations where energy levels are high.

That editable model matters more than it sounds. If the engineer keeps it and you don't, every future change means rebuilding the study from scratch. Owning the model turns the next revision into an update rather than a new project.