Condition of Maintenance: The Hidden Variable in Your Arc Flash Risk Assessment

In our NFPA 70E classes, I like to ask safety managers a version of one question: “Whose job is it to keep your arc flash risk assessment valid?” Almost everyone answers with an engineering title — whoever ran the incident energy study, or the consulting firm that did the arc flash analysis. In roughly twenty-five classes a year, I have never once heard someone answer “maintenance.”

That is the blind spot. Part of the answer to whether a task is even likely to produce an arc flash incident comes directly out of your maintenance program — not your engineering study, not your five-year review cycle. If maintenance and electrical safety are run as two separate conversations at your facility, there is a good chance your risk assessment is already resting on an assumption nobody is actively checking.

Table 130.5(C) Isn’t Just an Engineering Table — It’s a Maintenance Test

Table 130.5(C), Estimate of the Likelihood of Occurrence of an Arc Flash Incident, is central to how NFPA 70E determines whether a task requires additional protective measures. For a long list of tasks, the likelihood outcome depends entirely on one classification: is the equipment’s operating condition Normal, or Abnormal? That classification traces back to the same normal-operating-condition requirements in Section 110.2(B), Exception No. 1 — and the 2024 edition changed what it takes to satisfy them.

“Normal” now requires satisfying seven specific criteria, not six. All seven have to be true at the same time:

1. The equipment is properly installed in accordance with the manufacturer’s recommendations and applicable industry codes and standards.
2. The equipment is properly maintained in accordance with the manufacturer’s recommendations and applicable industry codes and standards.
3. The equipment is rated for the available fault current — new in the 2024 edition.
4. The equipment is used in accordance with instructions included in its listing and labeling and the manufacturer’s instructions.
5. Equipment doors are closed and secured.
6. Equipment covers are in place and secured.
7. There is no evidence of impending failure — arcing, overheating, loose or bound parts, visible damage, or deterioration.

Read that list again with a maintenance lens instead of an engineering one. Criterion two is a maintenance record, full stop. Criterion seven is exactly what a preventive maintenance program, an infrared scan, or a breaker test is designed to catch before it becomes a failure. Two of the seven gates that determine whether equipment even qualifies as “Normal” are maintenance outputs — not conclusions from an arc flash study.

The new third criterion is worth a second look too. An available fault current rating isn’t something maintenance checks on a PM ticket — it drifts when the utility upgrades a transformer or adds a substation upstream of you, which is exactly the kind of change a Management of Change process is built to catch. “Normal” condition now depends on your maintenance program and your MOC process both holding up their end.

Two Failure Modes, Not One

This is where it gets interesting for anyone running a program, because deferred maintenance doesn’t fail you one way. It fails you two ways, and they compound.

Failure Mode 1: The Severity Goes Up

Your incident energy analysis assumes a protective device that clears within its published time-current curve. A breaker that hasn’t been tested, cleaned, or exercised in years frequently clears slower than that curve assumes. Slower clearing time means more incident energy delivered to the worker — and the label on the equipment has no way of knowing that. It reflects the study’s assumptions, not the breaker’s actual condition.

Failure Mode 2: The Classification Changes

This one gets missed more often. If equipment fails even one of the seven “Normal” criteria, it is Abnormal — and Table 130.5(C) does not offer a “No” likelihood outcome for Abnormal equipment on the tasks it lists. Routine tasks that would otherwise carry no likelihood of an arc flash incident under Normal conditions — operating a breaker, opening a panelboard cover to reach dead-front devices — lose that exemption the moment the equipment can’t be verified as properly maintained. The task didn’t change. The maintenance history did.

The 2024 Edition Gave You an Audit Tool: Annex S

For years, “condition of maintenance” was a phrase every qualified person had to interpret for themselves — and interpretations varied by facility, by auditor, and by how the maintenance backlog happened to look that quarter. The 2024 edition narrowed that gap with a new informative annex, Annex S, Assessing the Condition of Maintenance, which points qualified workers toward actual sources of information to check: manufacturer maintenance recommendations, infrared inspection records, protective device test results, and documented maintenance history among them.

Annex S is informative, not mandatory — it won’t get you cited on its own. But it is the closest thing NFPA 70E has ever given you to an audit checklist for a question that used to be left entirely to judgment. If your program doesn’t already use it that way, that’s a gap worth closing before an auditor, or an incident investigation, closes it for you.

What This Should Change About How You Run Your Program

Most electrical safety programs and most maintenance programs are run by different people, reporting through different chains, tracked in different systems. NFPA 70E doesn’t care about your org chart. It ties the validity of your risk assessment directly to whether maintenance got done.

A few practical changes worth making:

• Loop maintenance backlog into your periodic risk assessment review. The five-year review cycle under Section 130.5(G) is a backstop, not a plan — and it is not a substitute for knowing whether your PM program is current.
• Use Annex S as a real checklist, not a reference nobody opens. Walk it against your actual maintenance records for your critical equipment, not just what’s convenient to review.
• Ask the uncomfortable question before your next audit does: if a qualified worker had to prove this equipment is “Normal” today, using all seven criteria, what would they actually find?
• Treat thermographic scans, breaker test reports, and PM completion data as electrical safety documentation — not just a reliability file that the safety program never sees.

The Bottom Line

The arc flash label on that equipment was accurate under an assumption: that the equipment behind it was properly maintained. Nobody rechecks that assumption automatically. A process does — the same discipline that a Management of Change process applies to system changes belongs here too, applied to maintenance status.

If you can’t say with confidence, right now, which of your critical equipment would pass all seven “Normal” criteria, that is your starting point — not the next incident energy study, and not the next scheduled audit.

For information on our NFPA 70E 2024-based training programs — available onsite at your facility or delivered live virtually — see Onsite Training and Virtual Training.