Fire drills and evacuation records: what you need to be able to show

Most buildings do fire drills. Rather fewer can produce, on request, a record that shows what was tested, what went wrong, and what was done about it. The gap between those two positions is where the awkward conversations happen, because a drill that was not recorded is — evidentially — a drill that did not happen.

This matters more than it used to. The direction of travel across fire safety regulation has been consistently toward documented evidence rather than assurance, and record-keeping duties on responsible persons have been tightened rather than relaxed. The specifics of what applies to a given building depend on its type, occupancy and height, and the current guidance is where those should be checked — but the general principle is not in doubt: if you cannot show it, you are relying on being believed.

A drill is a test, and a test that always passes is not testing anything

The most common failure in drill practice is not frequency. It is that the drill has been designed, often unconsciously, to succeed. Everyone knows it is happening. It is at a convenient time. The usual fire marshals are on shift. The route everyone uses is the route that is always available.

A drill run under those conditions tells you that people can leave a building they were expecting to leave, by a route they use every day, with the right staff present. That is worth knowing once. Repeating it annually for a decade produces ten records and almost no information.

A drill that tests something generally varies one thing: a blocked primary route so people have to use a secondary one, a time of day when the building is staffed differently, or a scenario where the person who normally takes charge is deliberately unavailable. The results are usually messier, which is precisely the point — the mess is the finding.

What the record needs to contain

A date, a time and a word like "satisfactory" is not a record in any useful sense. What makes a drill record defensible, and genuinely useful afterwards, is specificity.

• –Date, start time and the time the last person was accounted for — not just that the building was cleared.
• –What triggered the drill and how: which call point, or whether it was a silent or announced exercise.
• –Approximate numbers in the building, and whether that reflects a typical occupancy or an unusually quiet or busy period.
• –The scenario tested, including anything deliberately made unavailable.
• –Which routes were actually used, and which were not — people revert to the way they came in, and that is a finding.
• –Anyone requiring assistance to evacuate, and whether their personal arrangements worked in practice.
• –Problems observed: doors that did not release, alarm sounders that could not be heard in a particular area, confusion at the assembly point, people returning to collect belongings.
• –Who was present in which role, and whether the person who should have taken charge was available.
• –Actions arising, with an owner and a date.
• –Confirmation that each action was closed, recorded against the original finding rather than in a separate list.

That last point is where otherwise good practice tends to unravel. Findings get recorded diligently and then live in a document nobody revisits. The question an inspector or an insurer asks is not whether you found a problem — finding problems is evidence that the drill worked — it is what happened next. What record-keeping is expected of a responsible person covers the broader set; the drill-specific discipline is simply that a finding without a closure is an open admission.

Evacuation times are a comparison, not a target

Recording how long an evacuation took is useful. Treating it as a performance target is not, and it quietly distorts behaviour: a building that is being measured on clearance time will, over a few cycles, start running drills in conditions that produce good times.

The value in the number is comparative. Against the same building under similar occupancy, a time that drifts upward over successive drills is telling you something — about a route that has become congested, a change of layout, or a shift in who is in the building. Against a different building, or against a published benchmark, it tells you very little.

The things a drill tests that are not evacuation

Drills surface problems in adjacent systems, and these are frequently the most valuable findings because they are invisible the rest of the time.

Audibility is the classic one. Alarm coverage is designed and commissioned, then the building gets reconfigured — a partition goes in, a noisy plant item is installed, an area changes use. The drill is the only routine occasion on which anyone discovers that a particular corner cannot hear the sounder. Lighting on escape routes is similar: emergency lighting has its own testing regime, covered in how emergency lighting testing works, but a drill in winter darkness tests the thing those tests are a proxy for, which is whether people can actually see their way out.

Door hardware is a third. Doors on hold-open devices that do not release, final exit doors that stick, and doors that have been propped or locked for operational convenience all show up under drill conditions and rarely otherwise.

Who owns this

In a single-occupancy building the answer is usually clear. In a multi-occupied or managed building it frequently is not, and drills fall into the gap between a managing agent who considers them the occupier's business and occupiers who assume the building manages them.

This needs resolving explicitly rather than by custom, because the duty does not disappear simply because two parties each assumed the other held it. Who the responsible person is under the Fire Safety Order is the starting point, and where more than one party has duties in relation to the same premises, co-operation between them is part of what is expected rather than a courtesy. The practical output should be a written statement of who arranges drills, who attends, who records them and who holds the record — agreed in advance, not reconstructed after an incident.

Frequency, briefly

Expectations on how often drills should be held vary with the building, its occupancy and the risks identified in its fire risk assessment, and that assessment is the proper driver rather than a general rule of thumb — the current guidance for the premises type is where to check it. What is worth saying is that the interval should be justified by something written down, and that buildings with high turnover of occupants, shift working, sleeping risk or occupants needing assistance to evacuate will generally not be well served by the minimum anyone has heard quoted.

The broader duty context sits in what the Fire Safety Order actually requires in practice. Drills are one of the few parts of that framework where the evidence is generated by doing the thing rather than by commissioning someone to inspect it — which makes the quality of the record entirely your own responsibility.