Ask a building owner when their fire extinguishers were last inspected and you will usually get an answer, or at least a tag to point at. Ask when the doors on the exit routes were last tested and the question tends to land as a novelty.
Both are scheduled obligations. Only one of them has a service industry that shows up unprompted with a clipboard.
Why doors are life-safety equipment
An access control system spends almost all of its life doing the job you bought it for: keeping people out of places they should not be. It is judged on that, maintained on that, and quoted on that.
Then, rarely, it has to do the opposite job. In an evacuation the same doors must let everyone out, immediately, without a credential, without a phone call, and without anyone having to know how the system works. That second job is the one the codes care about, and it is the reason a locked door on an exit route is not purely a security device. It is part of the building's egress provision.
The awkward property of the second job is that it is invisible until it is needed. A door that has quietly stopped releasing on alarm looks exactly like a door that works. It swipes open normally every morning. Staff walk through it hundreds of times a week. Nothing about ordinary use exercises the failure mode, which is why the codes require the failure mode to be exercised deliberately, on a schedule.
What the Fire Code requires
BC's Fire Code sets testing intervals for doors in a means of egress at Article 2.7.2.1. In substance it establishes two cadences that matter to anyone running an access control system:
- Egress doors are to be tested monthly.
- Electromagnetic locks carry a twelve-month interval.
Two honest notes about how to use that. First, the Fire Code text sits behind the free-registration BC Codes viewer, so there is no stable public link we can send you to that opens directly at the clause — you or your consultant will need to sign in to read it in full. Second, and more importantly: confirm the current wording and the governing edition for your building before you build a programme around it. Code editions change and clause numbering has moved before. We have given you the Article number so you can find it; the version that governs your building is the one to work from, and your local authority having jurisdiction is the arbiter of that, not a vendor.
That caution is not throat-clearing. Our own article on electromagnetic locks exists partly because the 2024 Building Code renumbering caught out competitors who were still citing the old clauses. Getting the number right and the edition wrong is a real failure mode in this subject.
What "testing" actually means
This is where most well-intentioned compliance goes soft. A test is not a look. It is not a maintenance visit. It is a deliberate attempt to make the door do the thing it is supposed to do in an emergency, followed by a record that says it did.
For a door held by an electromagnetic lock, the honest test is uncomfortable and short: remove power and confirm the door releases. Not check the wiring. Not confirm the panel reports healthy. Actually drop the power and put a hand on the door.
The reason this matters more than it sounds is that from the inside, a mechanical lever and a fail-secure electrified lever look identical, and both feel identical when the system is working. The distinction only appears when power goes away. A building can run for years believing a door will release, and be wrong, and no one finds out from the panel — because a supervision signal tells you the device is answering, not that it is going to do the right thing when it stops being asked.
The same principle applies to the rest of the release path. If the door is meant to release on a fire alarm signal, the test has to include that signal arriving, not a technician pressing a button that simulates it downstream. If there is a request-to-exit device, a break-glass, or a delayed-egress function, each of those is a separate thing that either happens or does not. A door with three release mechanisms has three tests.
Who owns this
Almost always the building owner or operator, and almost always someone who does not know it.
The pattern we see is straightforward. Access control gets installed as a security project, sometimes by a security integrator, sometimes as part of a larger fit-out. The system is handed over working. The relationship afterwards is reactive — someone calls when a reader fails or a fob stops working. Nobody was assigned the recurring obligation, because at handover it did not feel like there was one.
Meanwhile the fire alarm system has an inspection contract, because that industry has spent decades making scheduled inspection the default. Access control has not, and the gap between the two is where this obligation lives unattended.
Three questions settle it for your building:
"Who tests our egress doors, and when did they last do it?" If the answer names your fire alarm contractor, confirm that their scope actually includes the door hardware and the electromagnetic locks, rather than the alarm system that signals them. Those are frequently different scopes and different trades, and the seam between them is exactly where the responsibility goes missing.
"Where are the records?" A test with no record is, for every practical purpose that matters later, a test that did not happen. Insurers and inspectors ask for the log.
"Which doors are on the list?" Buildings get modified. A door that was added, converted, or brought under access control during a tenant improvement three years ago may never have been added to anyone's schedule.
The failure that motivates all of this
The consequence of skipping the schedule is not a fine, in the first instance. It is a door.
A door that does not release when it should is a hazard at exactly the moment a building has the least capacity to deal with one. That is the actual reason the intervals exist, and it is worth keeping in front of the commercial framing, because the commercial framing — inspection cost, contractor scheduling, who pays — is what usually decides whether this gets done.
There is a secondary consequence worth naming for anyone weighing the cost. After an incident, the question of whether the building met its obligations is asked with the benefit of hindsight and documentation. A maintained testing log is a straightforward answer. Its absence is not neutral.
What good looks like
A short, boring, durable arrangement:
A list of every door under access control, marked for whether it sits on an exit route. A named party responsible for testing, with the scope written down and the electromagnetic locks explicitly included rather than assumed. Monthly and annual cadences held against the Code Article as it applies to your building. A log that records date, door, what was tested, and the result — including the failures, because a log with no failures in it over several years is usually a log of visits rather than tests. And a trigger that adds new doors to the list whenever the building changes.
None of that is expensive relative to the system it protects. It is mostly a decision about whose job it is, made once, and then written down.
If you do not currently know the answer to "when were our egress doors last tested", that is the finding. It is also, in our experience, the most common answer.
