The practical answer
Check hazard labels by confirming that each one names the equipment in front of you, carries every element the adopted edition requires, and traces to a dated analysis whose currency you can assess. Record presence, legibility, and condition separately, and route anything missing or doubtful to the named technical owner. Route each part of the label to the document that answers for it: the adopted edition of NFPA 70E for safe work practices, the adopted electrical code for installed marking, IEEE 1584-2018 for calculation, and the ANSI Z535 family for format. A labeling review never transcribes, converts, or infers a protective value, and it does not authorize work.
The method, examples, and sources below explain the scope and checks.
What this check covers, and what it does not#
An identification review can confirm that a hazard label exists, that it is legible, that it names the equipment in front of you, and that it agrees with the document that produced it. It cannot establish incident energy, select protective equipment, or authorize work on energized equipment. Those outputs belong to the electrical engineering analysis and to the employer's electrical safety program.
OSHA places the selection of safe work practices and the training of qualified persons on the employer rather than on the person reading the label. OSHA 29 CFR 1910.333 describes those work-practice obligations, and OSHA 29 CFR 1910.303 covers general electrical installation requirements including field marking. Neither makes a labeling survey a substitute for the analysis behind the label.
Use this guide to produce an exception list for the responsible technical owner. When the review finds a missing, unreadable, or doubtful label, the output is a dated record and a referral. It is not a replacement sticker.
Which document answers for which part of a hazard label#
A request to make the labels "NFPA 70E compliant" names one document and depends on four. They have different subjects, different publishers, and different silences. A review that collapses them onto one document produces findings nobody can act on, so route the question before checking anything.
| Question about a hazard label | Document whose subject covers it | What it still does not decide |
|---|---|---|
| Which safe work practices apply, and who is qualified to carry them out? | NFPA 70E, as adopted by the employer | Whether any particular task was performed safely |
| What marking must installed equipment carry? | NFPA 70, the National Electrical Code, as adopted and amended by the jurisdiction | Whether a given installation complies |
| How are incident energy and the arc flash boundary calculated? | IEEE 1584 | Which protective equipment to wear; its published scope excludes PPE recommendations |
| Who selects work practices and qualifies the worker? | OSHA 29 CFR 1910.333, with field marking in 1910.303 | The content of any particular label |
| What are the values for this panel? | Your own arc flash risk assessment, at its stated revision | Anything about equipment the study did not model |
| What signal word, color and layout should the label use? | ANSI Z535.4 for product safety labels and ANSI Z535.2 for facility safety signs, both current at 2023 | Which of the two your issuer applied, which is their decision and not yours to restyle |
| What marks an energy-isolating device while work is under way? | OSHA 29 CFR 1910.147, a different subpart with its own durability and legibility rules | Anything about the permanent hazard label on the same enclosure |
| Is the equipment maintained so the analysis still describes it? | NFPA 70B, current at 2026 | Whether any particular label is correct |
Two distinctions in that table cause most of the arguments. A consensus standard and a federal regulation are not interchangeable, and a calculation guide is not a protective-equipment selection. A labeling review can record which document a label claims to follow. It cannot move a question from one row to another.
Markings that look like a hazard label and are not#
One door in a data hall can carry five markings applied by five organizations under five different rules. A review that treats them as one class either raises findings against markings nobody in your organization owns, or misses the one that actually needed raising. Separate them before recording anything.
| Marking | What it is | Who owns the content | Where the rule comes from |
|---|---|---|---|
| Arc flash and shock hazard label | A field-applied warning carrying the output of an analysis | The electrical engineer of record, or the study provider | The adopted edition of NFPA 70E, with format conventions from the ANSI Z535 family |
| Equipment nameplate | The manufacturer's rating and identification plate | The equipment manufacturer | The product listing under which the equipment was evaluated |
| Field marking on installed equipment | Marking applied because the installation calls for it | The installer, accepted by the authority having jurisdiction | The adopted edition of NFPA 70, and 29 CFR 1910.303 |
| Accident-prevention tag | A temporary warning about a specific condition | Whoever hung it, for as long as it hangs | 29 CFR 1910.145, reached from 1910.335(b) for electrical work |
| Lockout or tagout device | A temporary marking on an energy-isolating device | The authorized employee who applied it | 29 CFR 1910.147 |
| Asset tag | An inventory identifier for something you own | Your own asset policy | Nothing external. This is the only row with no document behind it |
| Location or position label | A wayfinding and identification marking | Your identification program | The administration standard your project adopted |
Two rows cause most of the trouble. The nameplate is the manufacturer's statement about the equipment, applied under the listing it was evaluated against, so covering it, replacing it, or printing a tidier copy is not a decision a labeling program gets to make. The asset tag is the opposite problem: it is the only marking on the door with no external document behind it, which means nobody else will tell you what it should say or who issues the next value, and a program that never decides ends up with three conventions and no owner.
The temporary markings deserve their own note. A tagout device and a hazard label are both printed, both warn, and both end up on the same enclosure, but one is applied by the person doing the work and removed by them, and the other outlives everybody involved. Record them separately or the review will report a missing label that was never there and a stale label that was correct yesterday.
Editions: NFPA 70E-2027 is the current edition#
The edition a document is published at and the edition your work is governed by are two different facts. Most disputes about hazard labels come from treating them as one. Published editions are a matter of record and are listed below. Adopted editions are a matter of your own program and your jurisdiction, and this table cannot tell you those.
| Document | Current edition | What establishes it | Earlier editions still cited in the field |
|---|---|---|---|
| NFPA 70E, Standard for Electrical Safety in the Workplace | 2027 | Publisher standard-development page: issued April 16, 2026, effective May 6, 2026 | 2024, 2021, 2018 |
| NFPA 70, National Electrical Code | 2026 | Publisher code-development page; the 2029 revision cycle is open | 2023, 2020, 2017 |
| IEEE 1584, Guide for Performing Arc-Flash Hazard Calculations | 2018 | Publisher standard page: published November 30, 2018, superseding 1584-2002, listed as an active standard with project P1584 authorized for the next revision | 2002 |
| OSHA 29 CFR 1910 Subpart S | Not on an edition cycle | A federal regulation, changed by rulemaking rather than by revision cycle | Not applicable |
A published edition does not adopt itself. An employer's electrical safety program names the edition it follows, and a jurisdiction adopts an electrical code on its own schedule and with its own amendments, so a label referencing an earlier edition is a question to route rather than a defect to correct at the panel.
The calculation row is the one that surprises people. Because IEEE 1584-2018 supersedes the model published in 2002, two studies of the same equipment performed under the two models can report different values without either being an error. A labeling review records the study reference and the discrepancy; it does not decide which number is right.
Where hazard labels sit in a data center electrical path#
Most published writing about arc flash labeling pictures a factory motor control center. A data center's electrical path has its own shape and its own handover seams, and a review that walks it in order finds gaps in a repeatable sequence instead of wherever somebody happened to look. Walk it from the service entrance inward.
| Where you are standing | What a review usually finds | Who usually owns it | What makes it easy to miss |
|---|---|---|---|
| Utility service entrance and main switchboard | The highest available energy in the building, often behind the oldest label | The engineer of record, at the utility interface | It frequently sits in a room a facilities walk does not enter |
| Generator plant and paralleling switchgear | Markings from a different vendor package than the rest of the room | The generator vendor, or the engineer of record | It arrives as its own package with its own documents and its own dates |
| Automatic transfer switch | A marking whose values depend on which source is connected | Engineer of record | The condition it was analyzed under is not visible on the label |
| UPS, its maintenance bypass, and its battery system | Several markings within arm's reach on different assemblies | UPS vendor, plus the engineer of record | DC and battery hazards answer to different questions than the AC side |
| Distribution panelboards and remote power panels | The largest count in the building, and the most label churn | Facilities engineering | Quantity alone. A spot check under-samples the population that changes most |
| Busway and its tap-off units | Frequently no label, and no named owner either | Often unassigned | Installed by a different party, and the tap-off unit moves during its life |
| Rack PDU and in-rack distribution | An asset tag where a reviewer expected a hazard label | The IT asset owner | The nearest marking is an inventory tag, which invites the wrong comparison |
Whether a given assembly carries a hazard label at all is not a labeling decision. It follows from the adopted code, the employer's electrical safety program, and the tasks people actually perform on that equipment. What a labeling review contributes is the inventory: which equipment has a marking, which does not, and which has one nobody can trace. Route the question of whether a gap was required to be filled, rather than answering it at the panel.
Two patterns turn up in almost every walk. Equipment energized late in a build, such as a second UPS module or a bank of remote power panels added for a later hall, is the most likely to carry values from a study revision the rest of the room no longer runs on. And equipment installed by somebody other than the electrical contractor, which in a data center usually means the busway and the rack PDU, is the most likely to have no marking and no named owner, because it fell between two scopes of work at handover.
Identify who owns the label content before touching anything#
Every hazard label should trace to a controlled document: an arc flash risk assessment or incident energy analysis, a drawing revision, or an equipment submittal. Record the document name, revision, date, and accountable owner before evaluating the label itself. If the review cannot name that document, the finding is a records gap, and it is more important than any wording on the sticker.
Ask who is permitted to issue a replacement. In most organizations that authority sits with the electrical engineer of record, the facilities engineering group, or a contracted study provider, not with the team that owns the printer. Write that name into the review record so the exception has somewhere to go.
In a colocation or shared facility, ownership follows the equipment boundary rather than the room. Equipment owned and maintained by the provider carries labels the provider controls. Confirm the boundary in writing before recording an exception against something another organization maintains.
Whose label is it in a colocation space#
Ownership of a hazard label follows the equipment boundary, not the room. This is easy to say and hard to apply, because in a shared facility the equipment on either side of a boundary can be within reach of the same person. Settle it in writing before a review records a single exception, or the output is a list of findings addressed to an organization that will not accept them.
| Where the equipment sits | Who owns the equipment | Who owns the hazard label | What the other party can do |
|---|---|---|---|
| Provider power infrastructure upstream of the cage | The provider | The provider, from the provider's own analysis | Report an observation through the contract, and nothing else |
| The cage or suite distribution panel | Usually the provider, sometimes the tenant | Follows the equipment, and is worth confirming in writing rather than assuming | Ask which it is before the first walk, not after the first finding |
| Tenant rack PDU and in-rack distribution | The tenant | The tenant, if a marking is called for at all | Maintain its own records and its own exception list |
| Tenant IT equipment | The tenant | Generally no hazard label; the marking here is an asset tag | Keep the two classes separate in its own records |
The practical failure is not a dispute about who owns a panel. It is a review that records a finding, routes it to a name inside its own organization, closes it internally, and leaves a hazard label on somebody else's equipment untouched and unreported. Write the boundary and the reporting route into the review record itself, so an exception on the far side of it goes somewhere real.
A tenant is rarely able to resolve an exception on provider equipment, and should not try. What it can do is record the observation with a date and a photograph, report it through whatever channel the contract provides, and keep its own copy. That record is the tenant's evidence that it observed and reported, which is a different and more defensible thing than a record claiming the hazard was assessed.
Confirm the label belongs to the equipment in front of you#
A hazard label describes one piece of equipment at one point in a distribution system. Confirm that the equipment identifier printed on the label matches the identifier in your approved records for the physical unit you are standing at. This is the same crosswalk problem that appears between facility equipment names and operational names, and it fails the same way: two naming systems, one surface, and nobody reconciling them.
Where the label carries no equipment identifier at all, record that as a distinct finding. A correct set of values attached to an unidentifiable panel cannot be verified later, cannot be re-issued after a study update, and cannot be matched to a change record.
Check the label against the content list its own source requires#
Build the content check from the edition your program has adopted rather than from a remembered field list. NFPA 70E is the consensus document usually named for arc flash labeling in the United States. NFPA records the 2027 edition as current, issued April 16, 2026 and effective May 6, 2026, with 2024, 2021 and 2018 among the prior editions. Installation marking requirements sit in the adopted electrical code, published as NFPA 70, the National Electrical Code, whose current edition is 2026. Neither fact settles which edition applies to your equipment, because adoption is jurisdictional and a program names the edition it adopted. Record which edition and clause your check used, because that reference is what makes the result reproducible next year.
Structure the review so each element is confirmed separately rather than judged as one impression:
| Element to confirm | What the reviewer records | Why it is checked separately |
|---|---|---|
| Equipment identity | Identifier on the label and the matching record | A correct value on an unidentified panel cannot be re-issued or audited |
| Nominal system voltage | Value present and legible | Drives almost every downstream decision on the label |
| Arc flash boundary | Value present with its unit | An approach distance is a different quantity from a protective level |
| Protective level expression | Which form the label uses | Incident energy with working distance, a category, and a site-specific level are not interchangeable |
| Source analysis reference | Study name, revision, or date | Without it, currency cannot be assessed at all |
| Signal word and format | Wording and layout as issued | Format belongs to the issuing document, not to the reviewer |
Do not transcribe values between labels, translate a category into an energy figure, or infer a missing field from a similar panel. Each of those actions creates a new engineering statement that the reviewer is not authorized to make. Record the gap and stop.
Format and content come from different documents#
A hazard label is one surface assembled from four separate decisions, each made by somebody else. Reviewers get into trouble by treating a label they find ugly as a label they find wrong, and by treating a label that reads cleanly as one that must be current. Check the parts separately.
| Part of the label | Where it comes from | Who may change it |
|---|---|---|
| Signal word, color, and visual layout | The ANSI Z535 family: Z535.4 for product safety labels, Z535.2 for facility safety signs, both current at 2023 | The organization that issues the label, applying whichever it follows |
| Whether this equipment carries one | The adopted edition of NFPA 70 and the employer's electrical safety program | The engineer of record and the program owner, jointly |
| The values printed on it | The arc flash risk assessment, at a stated revision | Nobody except by re-issuing the analysis |
| Physical construction and permanence | The label manufacturer's evaluated marking and labeling system for that surface and exposure | The material decision, which is a procurement and engineering choice |
| Whether it is still right | The review interval in the adopted edition, plus your change records | Established by evidence, not by inspection of the sticker |
The Z535 row is the one most often collapsed. There are two standards in play because there are two situations: a label applied by a manufacturer to a product it sells, and a sign applied by a facility to its own environment. Field-applied hazard labels on facility switchgear have been issued under both, by different providers, in good faith. A reviewer records which convention the label appears to follow and moves on. Restyling somebody else's safety label to match a house template is not a tidy-up; it is an unauthorized change to a warning.
Protective equipment never appears in that table, and its absence is deliberate. Providing and maintaining protective equipment is an employer duty under 29 CFR 1910.132, with the electrical case at 1910.335, and the selection of safe work practices sits at 1910.333. The calculation guide is explicit in the same direction: IEEE 1584 states in its published scope that it does not include recommendations for personal protective equipment. Nothing about what to wear is an output of a labeling program.
Assess whether the label is still current#
A label can carry correct content and still be out of date. Treat currency as a separate question with its own evidence. The common triggers are a change to the upstream protective device, a change in available fault current, a transformer or service alteration, equipment replacement, and the elapsed review interval defined by the adopted edition and your own program.
Ask for the analysis date rather than assuming the label's age. Where the label shows no date and the study cannot be located, the correct outcome is an exception stating that currency could not be established, not a judgment that the label is wrong. Those are different findings and they route to different people.
Changes made by your own team are the easiest to miss, because they rarely feel like electrical modifications. A breaker swap, a feeder re-termination, or a panel schedule correction can all invalidate an existing analysis. Link the labeling exception to the change record that caused it wherever one exists.
Treat durability and legibility as their own finding#
A label whose content is correct and whose print has failed is still a failed label. Field-applied markings are expected to be permanent, legible, and suited to the environment where they are installed; the requirement belongs to the adopted electrical code, and the material evidence belongs to the manufacturer of the label stock.
Record the observed condition rather than a verdict: fading, lifting at an edge, surface contamination, a label applied over a previous label, or a position that has been covered by later equipment. Pass those observations to the material selection decision rather than solving them with whatever stock is loaded in the printer. Agency evidence such as UL recognition applies to the specific construction and surface the manufacturer identifies, not to a category of label.
Worked fictional case: the label outlived the equipment it described#
In fictional DC01 / H1, remote power panel RPP-2 carries a hazard label referencing study ARC-2021-04. The panel identifier on the label reads RPP-02 while the approved facility record and the door stencil read RPP-2. A change record from earlier in the year shows the main protective device was replaced during a capacity upgrade.
The review records three separate findings. First, an identifier mismatch between the label and the approved record, which is a crosswalk exception. Second, a currency exception, because a protective device change is a recognized trigger for re-evaluation and the label still references the earlier study revision. Third, no durability exception, because the print is intact and legible.
The reviewer issues no new label. The record goes to the named engineering owner with the panel identifier, the study reference, the change record number, and the date of observation. The identifier mismatch is fixed in the records and the door marking through the normal identification process, because that part is an identification problem. The protective-level content waits for the engineering owner. Every value in this example is fictional.
Record the exception rather than repairing the label#
Write the finding so that someone who was not present can act on it: equipment identifier, location, observed label content by element, what could not be confirmed, the document reference if one exists, the trigger that raised the currency question, photographs, reviewer, and date. Keep the physical label in place unless the responsible owner directs otherwise, because removing it destroys the evidence and leaves the equipment with no warning at all.
Track the exception to closure in the same system that holds other identification work, and record who accepted the closure. A labeling program can legitimately report that a hazard label is missing or unverifiable. It cannot report that the hazard has been assessed.
Acceptance checklist#
Frequently asked questions#
Can we print a replacement arc flash label ourselves?#
Only when the responsible technical owner supplies the content and authorizes the re-issue. Producing the physical label is a printing task; deciding what it says is an engineering output. Keep the two roles visible in the record so a re-issued label can be traced to the analysis revision it came from, not to the person who happened to run the printer.
The label has no date. Is it still valid?#
Record that currency could not be established, which is different from recording that the label is wrong. The content may be entirely correct. Without a study reference or date, neither the review interval nor the effect of later equipment changes can be assessed, so the finding is an evidence gap for the owner to close.
Who sets the review interval?#
The adopted edition of the consensus standard sets an outer limit, and your electrical safety program may set a shorter one. Record which document and interval the review applied. Equipment changes can require re-evaluation well before any interval elapses, so treat the interval as a maximum rather than a schedule.
What if the equipment identifier on the label does not exist in our records?#
Treat it as an identification exception first. Resolve the name through the normal crosswalk between facility records, operational records, and physical marking before anyone considers reprinting hazard content. A hazard label attached to an object your records cannot name is unverifiable regardless of the values it carries.
In a colocation space, whose label is it?#
Follow the equipment ownership and maintenance boundary stated in your agreement, not the physical room. Record the boundary in the review so exceptions are raised against the organization that can actually act. Where the boundary is unclear, that ambiguity is itself the finding worth escalating.
Does a labeling audit demonstrate electrical safety compliance?#
No. It demonstrates that labels were checked against a stated document on a stated date. Compliance with an electrical safety program involves analysis, training, work practices, and equipment condition that a label review does not observe. Report the checked scope precisely so the evidence is not read as more than it is.
What is the current edition of NFPA 70E?#
NFPA records the 2027 edition as current, issued April 16, 2026 and effective May 6, 2026. Many published references still describe the 2024 edition as current, which it was until the spring of 2026. The published edition is a separate fact from the one that governs your work, because an employer's electrical safety program names the edition it has adopted. Record the adopted edition in the review, since that is what makes a finding reproducible next year.
Does a new edition of NFPA 70E mean our labels are out of date?#
No. A newly published edition does not reach back and change what your program adopted, and it does not make correct labels wrong. What a new edition raises is an applicability review: someone decides whether to adopt it, on what schedule, and what that changes. Relabeling can be an outcome of that decision. It is never the first step.
Is it the label or the arc flash study that gets reviewed on an interval?#
The review interval in the adopted edition attaches to the risk assessment rather than to the physical label, and the distinction matters in both directions. A label that still matches a current assessment is not out of date because it is old. A freshly printed label reproducing a superseded study is out of date the day it goes up. Treat the age of a label as a reason to go and find the assessment, not as a finding in itself.
Does IEEE 1584 say which protective equipment to wear?#
No. Its published scope states that the guide does not include recommendations for personal protective equipment. What it provides is the calculation models for incident energy and the arc flash boundary. Selecting protective equipment belongs to the employer's electrical safety program, informed by that analysis. A label citing the calculation guide as the authority for a protective level has named the wrong document.
Why do two labels on the same equipment show different incident energy?#
Ordinary reasons include a change to the upstream protective device, a change in available fault current, and a different working distance. One more is the calculation model itself: IEEE 1584-2018 superseded the 2002 model, so two studies of the same equipment can legitimately differ according to which model each was performed under. Record both study references and raise the difference. The review does not adjudicate which value is correct.
Is an arc flash label the same as an equipment nameplate?#
No, and the distinction decides what you are allowed to do. A nameplate is the manufacturer's rating and identification plate, applied under the listing the equipment was evaluated against. A hazard label is field-applied and carries the output of an analysis performed on your installation. A review records both, but a nameplate finding routes to the equipment owner and the manufacturer, never to the team with the printer.
Can we cover or replace a manufacturer's nameplate with our own label?#
Treat that as out of scope for a labeling program. The nameplate is part of how the equipment was evaluated and identified, so covering it, relocating it, or reprinting a tidier copy changes something your program does not own. Where a nameplate is illegible or obscured, the finding is an exception addressed to the equipment owner, with a photograph, exactly like any other observation the review is not authorized to fix.
Which ANSI Z535 standard applies to a label on facility switchgear?#
Both are current at 2023 and they answer different situations. Z535.4 covers product safety signs and labels, meaning those a manufacturer applies to a product. Z535.2 covers environmental and facility safety signs, meaning those a facility applies to its own environment. Field-applied hazard labels on facility equipment have been issued under both conventions in good faith by different providers. A review records which convention a label appears to follow. It does not restyle somebody else's safety label to match a house template.
Does every panel in the data hall need an arc flash label?#
That is not a labeling decision and a labeling review should not answer it. Whether a given assembly carries a marking follows from the adopted code, the employer's electrical safety program, and the tasks people actually perform on that equipment. What the review contributes is the inventory: which equipment carries a marking, which does not, and which carries one nobody can trace to a document. The question of whether a gap was required to be filled goes to the named technical owner.
Put the decision into a record
EDITABLE PLANNING DOCUMENT
Electrical hazard label verification record
Record what a hazard-label review observed, which document edition it used, and which exceptions were routed to the responsible technical owner.
Sources and applicability
- NFPA: NFPA 70E standard development and editions ↗
Publisher standard-development page. Establishes 2027 as the current edition, issued April 16, 2026 and effective May 6, 2026, and lists prior editions back to 1979. The clause-level content list is in the document itself and is not reproduced here.
Read scope and linked guides - NFPA: NFPA 70, National Electrical Code ↗
Publisher code-development page for the installation code that carries field marking requirements. Establishes 2026 as the current edition, with the 2029 revision cycle open. Adoption, edition, and amendments vary by jurisdiction; confirm locally rather than assuming the latest edition applies.
Read scope and linked guides - OSHA: 29 CFR 1910.333 ↗
Regulatory text placing selection of safe work practices on the employer. Cited to establish that a labeling survey does not authorize work, not as a labeling content specification.
Read scope and linked guides - OSHA: 29 CFR 1910.303 ↗
General electrical installation requirements including field marking of equipment. Used for scope context only.
Read scope and linked guides - OSHA: 29 CFR 1910.145 ↗
Specifications for accident prevention signs and tags. Supports the separation of signal-word format questions from technical content questions.
Read scope and linked guides - IEEE SA: IEEE 1584-2018 standard page ↗
Publisher standard page. Establishes IEEE 1584-2018 as an active standard, published November 30, 2018, superseding 1584-2002, with project P1584 authorized for the next revision. Its published scope states the guide does not include recommendations for personal protective equipment. No calculation model or value is reproduced here.
Read scope and linked guides - OSHA: 29 CFR 1910.335 ↗
Safeguards for personnel protection, in Subpart S. Cited for two things it settles: providing protective equipment is an employer duty, and safety signs and tags for electrical hazards are reached through 1910.145. Neither is a label content specification.
Read scope and linked guides - OSHA: 29 CFR 1910.147 ↗
The control of hazardous energy, lockout and tagout, in Subpart J. Cited to separate a temporary marking on an energy-isolating device from a permanent hazard label on the same enclosure. The two answer to different rules and have different owners.
Read scope and linked guides - OSHA: 29 CFR 1910.132 ↗
General requirements for personal protective equipment, in Subpart I. Cited to place the provision and maintenance of protective equipment with the employer, which is why no protective-equipment decision is an output of a labeling review.
Read scope and linked guides - ANSI Z535.4-2023, product safety signs and labels ↗
Publisher listing for the current edition of the product safety label standard, developed by NEMA. Cited to place signal word, color and layout with their own document. No specification, color value or layout rule is reproduced here.
Read scope and linked guides - ANSI Z535.2-2023, environmental and facility safety signs ↗
Publisher listing for the current edition of the facility safety sign standard, developed by NEMA. Cited alongside Z535.4 because a field-applied label on facility equipment can reasonably be issued under either, and a review records which rather than restyling it.
Read scope and linked guides - NFPA: NFPA 70B standard development and editions ↗
Publisher standard-development page for electrical equipment maintenance. Establishes the current edition as 2026, issued November 19, 2025 and effective December 9, 2025, and lists the prior editions. Cited because maintenance is what keeps the condition an analysis described true, not because it specifies label content.
Read scope and linked guides - UL Solutions: marking and labeling systems program ↗
Publisher program page for label constructions evaluated for permanence of marking on a stated surface and exposure, under UL 969. Cited to separate an evaluated construction from a category of label. No test result, rating or product claim is reproduced here.
Read scope and linked guides