One Emergency Action Plan Template, Fifty Buildings: Where the Gap Actually Opens

A corporate safety team that manages forty locations has, in most cases, exactly one emergency action plan document. It has the company logo, a legal review behind it, a version number, and a distribution list. It gets updated on a schedule, pushed out as a PDF, and acknowledged by site managers who file it and move on. From a document-control standpoint, that looks like a mature program. From the standpoint of 29 CFR 1910.38, it may not be a plan at all — it may be a template that has never been finished at any of the forty sites it was sent to.

What the standard actually asks for

1910.38 requires an emergency action plan that is specific to the workplace: procedures for reporting a fire or other emergency, evacuation procedures and exit route assignments, procedures for employees who perform or shut down critical operations before evacuating, procedures to account for all employees after an evacuation is completed, rescue and medical duties for any employees assigned to perform them, and the name or job title of individuals who can be contacted for further information. Every one of those elements is a physical, site-specific fact. Exit route assignments depend on which doors exist and where they lead in that specific building. Accounting for employees after evacuation depends on where that site's assembly point actually is, and who is responsible for taking the headcount. None of that transfers from one address to another, no matter how well-written the corporate template is.

A document that lists these categories in the abstract — "employees will proceed to the designated assembly area" — without naming which door, which parking lot, and which supervisor performs the count at each site has satisfied the outline of 1910.38 without satisfying the requirement. The standard is written at the level of "what happens in this building," not "what our company's philosophy on evacuation is."

Why the gap is structural, not a compliance failure at any one site

No single site manager decided to under-comply. The gap opens because the organization built its emergency planning process around the wrong unit. Corporate legal and safety teams reasonably want one reviewed, defensible document rather than fifty independently drafted ones — the version-control and liability logic of a single template is sound. But OSHA's coverage obligation runs to the establishment, not to the corporate entity that owns it, and an establishment-level requirement satisfied by a company-level document is a mismatch that tends to stay invisible until an inspection or, worse, an actual emergency reveals it.

The sites where this shows up hardest are the ones added after the template was written: an acquired facility with a floor plan nobody at headquarters has walked, a leased space in a shared building where the assembly point has to be coordinated with other tenants, a site that added a mezzanine or reconfigured a warehouse floor without anyone connecting that physical change to the emergency plan sitting in a binder. The template doesn't know any of that happened, and nothing in a standard document distribution and acknowledgment process forces the question.

What "site-specific" costs, and what it doesn't

Making a corporate EAP framework actually satisfy 1910.38 at every location doesn't mean writing fifty unrelated documents from scratch. The parts of the plan that are genuinely company-wide — reporting procedures, the chain of authority, training requirements, the format itself — can and should stay standardized. What has to be completed locally is a short, specific annex per site: the actual exit routes for that floor plan, the actual assembly point and who counts heads there, the names of the people at that location who hold rescue, medical, or shutdown duties, and confirmation that alarm and notification systems at that site work the way the plan assumes they do.

That annex is also the piece most likely to go stale, which is why a completed EAP at the moment of the acquisition or the lease signing isn't the finish line. A site that adds a new production line, closes off an exit for renovation, or relocates its break room assembly point has changed the facts the plan depends on, and the update cycle needs to be tied to those physical changes at the site level, not to a corporate calendar that revisits the template annually regardless of what's changed underneath it.

Drills expose what the document can't

A written annex can look complete and still fail the first time it's tested, because 1910.38(e) requires reviewing the plan with employees when it is first developed, when an employee's responsibilities under it change, or when the plan itself changes — not simply distributing it once. A site that has never actually walked its evacuation route or run a headcount at the assembly point will discover, usually during the drill itself, that the door listed as a primary exit is now blocked by a pallet rack, or that the assigned counter left the company eight months ago and nobody reassigned the duty. Those are the failures a document review will never catch, because the document still reads correctly. Only a drill, conducted at that specific site, tests whether the physical world still matches the paper.

At scale, this argues for treating the drill schedule itself as the audit mechanism rather than a separate compliance task layered on top of the plan. A site that has run a documented drill in the past year has effectively re-verified its own annex. A site that hasn't is an unknown, regardless of how recently the written plan was distributed or acknowledged.

Contractors, temporary staff, and shared buildings complicate the count

The accounting-for-employees requirement gets harder, not easier, at sites that regularly host contractors, temporary staffing, or tenants who don't appear on the site's own payroll roster. A plan that only accounts for direct employees has a gap the moment a contracted crew is on-site for a shutdown, a staffing agency has workers on the floor, or the facility shares a building with another company's employees who evacuate through a jointly used exit. The plan needs to specify how those individuals get counted and who is responsible for it, because "someone will account for everyone" is not a procedure a drill can verify.

Multi-tenant and leased sites add a coordination layer that corporate templates rarely anticipate: the assembly point may need to be negotiated with a landlord or other tenants, and the responsibility for triggering an evacuation alarm may not sit entirely within the organization's own chain of command. None of that is resolved by a better-written corporate document. It's resolved by someone confirming, site by site, what the actual building and its other occupants require.

The visibility problem behind the compliance problem

The harder question for an organization running dozens or hundreds of sites is not whether the requirement is understood — it usually is, at the corporate level. It's whether anyone can currently answer, without calling around, which sites have a completed site-specific annex, which are still running on the unfinished template, and which haven't been revisited since a physical change made their existing annex wrong. That's a visibility question before it's a drafting question, and it's the one most likely to surface only when an OSHA inspector asks a site manager to walk the evacuation route and the manager can't confirm it matches what's on file.

An organization that can produce, site by site, a completed annex with a current review date has a materially different position than one that can produce a single well-written corporate template and an assumption that local managers filled in the blanks. The second position is common, defensible-looking on paper, and the one that tends not to survive contact with either an inspection or an actual evacuation.

If your organization is trying to get a real answer to which sites have a finished, current emergency action plan and which are still running on an unfinished template, that's the kind of cross-site visibility gap worth raising with your safety and compliance leadership — the fix is rarely more policy at the top, it's confirming what's actually true at each site underneath it.