Skip to content

Field guide

How Often Is an Arc Flash Study Required? The 5-Year Rule

Reviewed July 19, 2026 Standards-linked editorial Review policy

NFPA 70E requires the arc-flash risk assessment to be reviewed at intervals not exceeding five years. That is the maximum interval, not a recommended schedule — and it is only half the rule. The other half: the assessment must be updated when the system changes, whenever that happens. The five-year language was restated in the 70E-2024 cycle coverage of NEC 2026 label changes, and the same cadence is standard practice under Canada's CSA Z462.

Last reviewed: July 18, 2026.

The two clocks every label runs on

Clock 1 — the five-year review. Every assessment gets reviewed at least once per five years, even in a facility where nothing has moved. Dust, corrosion, undocumented swaps by a previous contractor, and utility-side changes all accumulate silently; the review is how you find out whether the numbers on the label still describe the equipment.

Clock 2 — change triggers. Any modification that can move fault current, clearing time, or configuration restarts the assessment for the affected equipment immediately:

  • Service upgrade or utility change — new transformer, larger service, or the utility revising its available fault current figure.
  • Protective-device swap — a different breaker or fuse class changes the clearing time, and clearing time is the strongest lever on incident energy.
  • Equipment replacement or modification — new switchgear, MCC sections, added panels, bus changes.
  • New loads with motor contribution — large motors feed fault current for the first cycles of a fault.
  • An incident or near-miss — the post-event review under your 70E program should include re-checking the assessment that the work was based on.

If a change trigger fires in year two, you do not wait three more years. The five-year clock is a backstop for the things nobody wrote down.

Review vs restudy — the honest distinction

These words get used interchangeably, and the difference is money:

  • A review verifies that nothing material changed: walk the one-line, confirm protective devices and settings match the record, confirm the utility fault-current figure is still current, and document the check. If everything matches, the existing assessment stands — no new calculation, and the documentation of the review itself is the deliverable.
  • A restudy re-runs the numbers. New field data, new calculation, new labels. Required when a change trigger fired, when the original inputs were estimates that can now be measured, or when the method itself moved (for example, a study run under IEEE 1584-2002 that has never been re-run under the 2018 model).

Facilities with good records review cheaply and restudy rarely. Facilities with poor records often have to restudy at what should have been a review, because nobody can confirm what changed. The cheapest five-year cycle is the one where every panel change gets logged when it happens.

Where NEC 2026 tightens the screw

NEC 2026 §110.16 requires the label to show the date the assessment was completed. That turns the five-year cycle from an internal policy into a field-visible fact: any inspector, insurer, or customer reading the label can see the assessment's age at a glance. A 2019 assessment date on a 2027 label is a finding waiting to be written up. Adoption is state-by-state — no states had adopted NEC 2026 at research time, Washington targets December 31, 2026 — but the 70E five-year rule applies today regardless of your state's NEC edition.

What this looks like in practice

  1. Put the assessment date on every label (required) and the review due date in your safety program's calendar.
  2. Log every change trigger as it happens — a one-line entry per panel is enough.
  3. At the review: confirm inputs, re-run only the panels whose inputs moved, reprint those labels.
  4. Keep the old calculation records. The label ID and stored inputs are what let a reviewer confirm "same system, same numbers" without a site visit.

The record-keeping half of this is what our tool stores by default: every calculation keeps its inputs, method, engine version, and a deterministic label ID, so a review is a comparison, not an excavation. The Pro tier adds five-year reassessment reminders against that stored record. Run or re-run a panel in the free arc flash calculator, and see arc flash label requirements for the date field NEC 2026 made mandatory. For budgeting the full-facility version, arc flash study cost breaks down the numbers.

Direct answers

Frequently asked questions

How often must an arc flash study be reviewed?

NFPA 70E sets the maximum review interval at five years — the assessment behind every label must be reviewed at least that often. The same cadence is standard practice under CSA Z462.

What triggers an arc flash study update before five years?

Any change that moves the inputs: a service upgrade, a utility fault-current revision, a protective-device swap, equipment replacement or modification, and large new motor loads. The five-year interval is a backstop, not a schedule.

Is a review the same thing as a new study?

No. A review verifies that nothing material changed and documents the check — the existing assessment stands. A restudy re-runs the numbers with new data. Good change records keep reviews cheap and restudies rare.

Does NEC 2026 change the five-year cycle?

Not the cycle — its visibility. NEC 2026 §110.16 requires the assessment completion date on the label, so the assessment's age is visible to any inspector reading the label.

Source trail

Sources

Standards references identify the applicable document or section where possible. Standards text may require licensed access. Report a factual issue through the process on our corrections page.