Arc Flash Boundary and PPE Selection Before Panel Work

Purpose

This settles two questions before a door or deadfront comes off: how far back everyone who is not working stands, and what the person at the door is wearing. It exists because an arc flash label is not a property of the panel. It is a dated calculation made under assumed source impedance, assumed clearing time and one specific working distance, and it stops being true quietly when the utility changes a transformer, when a breaker is swapped for a different frame, or when the task pulls the technician closer than the distance the number was computed at.

A shop that reads the category off the label and dresses to it is right most of the time. The failures cluster where the label is stale or where the task geometry differs from the study's, and in those cases the gear is not marginal, it is wrong by a wide margin. Neither failure looks like anything from the front of the panel, which is why it gets a written check rather than a habit.

Scope

Covers the pre-entry assessment for opening any enclosure at 50 V AC or more where energized parts may be exposed: panelboards, switchboards, motor control centers, meter mains, disconnects and junction boxes.

Does not cover the isolation sequence, which the lockout SOP owns, or the decision to work energized at all, which the energized work permit SOP owns. This procedure runs before both. Establishing an electrically safe work condition is itself a task performed inside the boundary, so the PPE selected here is worn while proving the equipment dead.

Roles and responsibilities

Procedure

1. Confirm the entering tech is qualified for this equipment class and task, and physically stop everyone else at the room door. Acceptance: they can state the voltage class, the interrupting method and this task's hazards, with documented 70E training inside the shop's retraining cycle. Wrong looks like a helper following in to hold the light. Stop rule: an unqualified person does not cross the doorway, and where there is no door, mark a line with cones first. Hazard: an arc flash injures bystanders well outside arm's reach, and a bystander is not dressed for it.

2. Find every label and read all of it, not just the largest number. Acceptance: you have written down incident energy in cal/cm2, the working distance it was computed at, the arc flash boundary, nominal voltage, the shock approach boundaries and the study date. NFPA 70E-2021, 130.5(H) sets the content of that field-applied label, while NEC 110.16 requires the arc flash warning marking itself on equipment likely to be examined while energized. Wrong looks like a sticker carrying a PPE category and nothing else. Stop rule: a label missing the working distance or the date is treated as no label and routed to step 4. Hazard: a category alone tells you nothing about whether the study's parameters still apply.

3. Check the label's currency. Acceptance: the study date is within 5 years, the review interval NFPA 70E-2021 sets in 130.5 for the arc flash risk assessment, and no major modification has occurred since. Wrong looks like a label from a study predating everyone currently at the shop. Stop rule: past 5 years, or after any modification, the label does not govern and you go to step 4. Hazard: the drift runs dangerous as often as safe, because a utility upgrade raises available fault current while a breaker change can lengthen clearing time.

4. Where the label is absent, stale or contradicted, choose between an updated study and de-energizing. Acceptance: either the engineer of record supplies a current incident energy figure, or the job is re-scoped to an electrically safe work condition established from a source outside the questioned equipment. Wrong looks like defaulting to the highest arc rating on the truck and calling it covered. Stop rule: an unknown incident energy is a scope problem, not a PPE problem; do not proceed energized. Hazard: guessed protection fails at the moment it is tested, and by then the tech has committed his hands.

5. Confirm the label's assumptions still describe the equipment in front of you. Acceptance: the utility transformer, the upstream device frame and setting, and the conductor run all match the study report, checked against the single line. Wrong looks like a 400 A main in the field where the report says 225 A, or a new pad transformer in the yard. Stop rule: any mismatch invalidates the label; return to step 4. Hazard: lowering upstream instantaneous settings lowers incident energy and raising them increases it, so an unlogged maintenance change moves the number both ways.

6. Mark the arc flash boundary on the floor and clear it. Acceptance: the boundary distance from the label is measured with a tape from the enclosure face, marked with tape or cones, and nothing flammable, no ladder and no second worker is inside it. That boundary is the distance at which incident energy falls to 1.2 cal/cm2, where bare skin takes a second-degree burn. Wrong looks like eyeballing it in a crowded electrical room. Stop rule: if the room cannot hold the boundary, clear the adjacent space or de-energize. Hazard: at the boundary itself an unprotected face sits at the burn threshold, not below it.

7. Measure the working distance the task actually requires and compare it to the label's. Acceptance: the distance from the tech's chest to the prospective arc source, taken at the deepest reach the task needs, is equal to or greater than the working distance printed on the label. Wrong looks like a label computed at 18 in for a task that puts the chest at 12 in to reach a lug in the bottom gutter. Stop rule: closer than the label's distance and the label's value does not apply; get the study's value at the actual distance, restore the distance with an insulated extension tool, or de-energize. Hazard: incident energy climbs steeply as distance falls, and no remembered exponent is accurate enough to scale it by eye.

8. Select arc-rated clothing whose arc rating meets or exceeds the incident energy at that distance. Acceptance: the garment label states an arc rating in cal/cm2 at or above the figure, the ensemble covers head, face, neck, torso, arms, hands, legs and feet, and every layer underneath is arc-rated or natural fiber. Wrong looks like an 8.0 cal/cm2 coverall against an 8.4 cal/cm2 exposure, or a polyester undershirt. Stop rule: a shortfall of any size sends the job back to step 4. Hazard: meltable synthetics under an arc-rated layer melt to skin and turn a survivable exposure into a graft.

9. Select shock PPE and check its test date before it goes on. Acceptance: rubber insulating gloves are class-rated above the nominal voltage, Class 00 to 500 V and Class 0 to 1,000 V under ASTM D120, leather protectors are on, and the electrical test date is inside the 6-month in-service interval OSHA sets at 1910.137(c)(2). Wrong looks like a stamped date nobody has read since the gloves went in the bag. Stop rule: an out-of-date or air-test-failing glove is bagged for return, not worn once more. Hazard: gloves fail at pinholes you cannot see, which is why the roll-up air test happens at the panel rather than in memory.

10. Run the final go or no-go with a second person present before the first fastener turns. Acceptance: every line on the pre-entry record carries a value rather than a check mark, and the attendant knows the limited approach boundary of 3 ft 6 in for exposed fixed parts at 151 V to 750 V AC, the restricted approach boundary of 1 ft 0 in, where the upstream disconnect is, and how to call for help. Wrong looks like a solo tech in an empty building with nobody aware he is in the panel. Stop rule: no attendant, no entry above 50 V. Hazard: an arc flash victim cannot self-rescue, and whoever reaches him has to know not to make contact while the source is still live.

The record this produces

One pre-entry record per equipment per visit, on the ticket rather than in a truck binder. Fields: equipment ID and location, date and time, tech name and qualification basis, label incident energy with units and working distance, label date, arc flash boundary as measured, task working distance as measured, garment arc rating, glove class and test date, attendant name, and the go or no-go with its reason.

Two people read it later. The estimator reads it when the site calls again, because a label found stale once will be stale again and the next scope should carry the study request. The supervising master reads the no-go entries monthly, because a site producing repeated no-gos needs a study quoted, not a braver technician. A record kept only for the jobs that went ahead loses the more useful half of the data.

Worked pass: 208Y/120 V retail distribution panel, adding a feeder tap

Ticket is to land a new 60 A feeder on spare lugs in a 208Y/120 V panel behind a strip retail unit.

Step 2 finds a label reading 8.0 cal/cm2 at an 18 in working distance, arc flash boundary 47 in, study dated 3 years ago. Step 3 passes on currency, since 3 years is inside the 5-year interval and the facility contact confirms no equipment change. Step 5 matches the single line to the field: 225 A main, as the report says. Step 6 measures 47 in from the enclosure face, which reaches past the aisle into a stock area, so the manager moves two pallets and the tech cones the line.

Step 7 fails. The spare lugs sit at the bottom of the gutter, and with the deadfront off the tech's chest is about 12 in from the prospective arc source, not 18 in. The 8.0 cal/cm2 figure was computed at 18 in and says nothing about 12 in. The stop rule fires: do not scale it by eye, and do not assume the truck's 12 cal/cm2 suit covers the gap, because the value at 12 in is unknown rather than known to be under 12.

Two resolutions were weighed. The engineer of record could model the incident energy at 12 in from the same study, which takes a day. Or an insulated extension tool could restore 18 in, which the lug geometry does not allow, since the conductor has to be seated and torqued by hand.

So the job was re-scoped. The customer agreed to a Sunday shutdown, the tap was landed in an electrically safe work condition, and the PPE worn for the isolation and verification was selected from the same label at the 18 in distance the deadfront-on switching operation actually uses.

That visit's record reads no-go, reason "task working distance 12 in against label working distance 18 in." That line is what let the estimator quote the Sunday window instead of arguing about it, and it is exactly what would have been missing had step 7 been marked with a tick.

References

• NFPA 70E-2021, Article 130, including the arc flash risk assessment and its 5-year review interval at 130.5 and the PPE requirements at 130.7, in the edition your electrical safety program has adopted
• 29 CFR 1910.132 (PPE hazard assessment) and 29 CFR 1910.335 (PPE and safeguards for electrical work), which bind the employer for its own employees
• 29 CFR 1910.137 (electrical protective equipment), including in-service test intervals for rubber insulating gloves
• NEC (NFPA 70) 110.16 arc flash labeling and 110.24 available fault current marking on service equipment
• ASTM D120 (rubber insulating gloves), ASTM F1506 (arc-rated clothing), IEEE 1584 (incident energy calculation)
• See related: the energized work permit SOP, which owns the justification test, and the lockout SOP, which owns proving dead