Testing an emergency exit light is not complicated, which is why it is so often done badly. There are two procedures, one short and one long, and the difference between doing them and being able to show you did them is entirely in what gets written down while you are standing under the fitting. This is the practical version: what you press, what you watch, what you record, and the three mistakes that turn a completed round into an unprovable one.
The monthly function test, fitting by fitting
Work a floor at a time with the device list in front of you. Operate the test switch on the fitting, or drop the local lighting circuit if the fittings have no switch, and watch: the lamp should come on immediately and stay on while the supply is off. Look at the lens for damage and discolouration and at the charge indicator for a healthy state. Restore the supply and confirm the charge indicator returns. Record the fitting identifier, the date, that it was a functional test, the result, and your name, then move on. Most fittings take a minute or two.
The annual duration test, and the ninety minutes you have to plan for
The annual test runs the fittings on battery for the full rated duration, and the guidance cited below states that a minimum of 90 minutes of illumination is required. Plan the visit around that: isolate the supply, note the start time, walk the site at intervals rather than sitting in the van, and at the ninety-minute mark record which fittings are still lit at an adequate level and which are not. Any fitting that failed early needs its failure time noted, because seventy minutes and ten minutes are different problems with different remedies.
The three mistakes that make a good round unprovable
First, testing everything and recording a site-level pass, which cannot answer a question about one fitting. Second, doing the annual test on a Friday afternoon and restoring the supply before ninety minutes because the day ended, which is a test that proves nothing and reads as a pass. Third, writing the results on a sheet that lives per visit rather than per fitting, so a fitting's history is a folder search. The first two are discipline; the third is why device-level records exist.
Before you plan the year, size it
Ninety minutes per site is wall-clock time somebody is standing there for, and across a round of sites that is the biggest single block of emergency lighting work in the year. The free emergency lighting test log on this site takes your fittings, your self-test share, your travel time and your hourly cost and returns the hours and the dollars the whole year takes, monthly rounds and annual discharge together, so the schedule is a decision rather than a surprise.
Questions people ask about how to test emergency exit lights
Do I need to isolate the whole board for the annual test?
You need the fittings running on battery, and how you achieve that depends on the installation: a test key switch, a local circuit isolation, or a central inverter's own test mode. Isolating more than you need to is how a duration test turns into an incident, so agree the method with whoever is responsible for the building's supply before the day.
What if a fitting fails at sixty minutes?
Record the failure time, not just the failure. Sixty minutes usually means a tired battery and a battery change; a fitting that never lights at all is more likely a lamp, a driver or a charging fault. Writing the minute down is what lets next year's technician see whether the replacement fixed it or the fitting is the problem.
Can one person test a large site alone?
For the monthly walk, usually yes. For the annual duration test on a large site, the constraint is not the labour but the geography: the fittings have to be observed at the end of the ninety minutes, and a single person cannot be on four floors at once at minute ninety. Either split the site across visits or take a second pair of eyes, and record which fittings each person read.
How do I test an exit sign as opposed to an emergency light?
The same way, with one addition. An illuminated exit sign with battery backup gets the function test and the duration test like any fitting; a sign that is photoluminescent or externally lit gets a visual check that it is legible from the route and that its light source works. Record which kind each sign is, because the test that applies follows from it.