Most atrium maintenance schedules fail for the same reason: they were written by someone who wasn’t going to have to follow them.
They’re thorough. They list every component, assign every task a frequency, and they sit in a folder nobody opens because the first quarter’s actions needed three contractors, a permit and a closed lobby. By year two the schedule has been quietly replaced by waiting for something to leak.
A schedule that gets followed looks different. It’s shorter, it’s built around how many times a year anyone actually goes up there, and it separates what the in-house team can do from what needs a specialist.
Start with the access mobilisations, not the tasks
This is the part most schedules get backwards. They list what needs doing and work out the access afterwards, which is how you end up with four separate visits to height in a year, each one carrying its own rigging, permits and disruption.
Access is the expensive half of any high-level glazing job. The glass work is often a fraction of the bill.
So decide first how many times a year someone goes up with close access to the glass. For most urban commercial atria that’s twice. Then load everything onto those visits.
A visit that cleans the glass, checks the perimeter seals, assesses the gaskets, looks at the fixings and clears the gutters and outlets costs barely more than a visit that only cleans, and it gets you the condition of the roof as a by-product. Running that as two separate contracts means paying for the access twice. The argument for doing it this way round is the same one behind planned atrium maintenance beating reactive repairs, and it mostly comes down to mobilisation cost.
What the in-house team can genuinely do
More than people assume, as long as nothing needs height.
From the floor, a quarterly walk looking up tells you a lot. Misted units are visible. Staining along glazing bars is visible. New cracks are visible if someone has a baseline to compare against. Water marks on walls and ceilings below the roof line are the clearest signal there is, and nobody needs a rope to spot one.
After any storm, a look up and a look at the floor. Fragments, drips, new marks.
Internal glass and the drainage at low level are both reachable. So are the ground-level glazed elements that frame the space, and keeping an eye on those alongside the roof is sensible since glass partitions in a small or busy space take more knocks than overhead glazing ever does.
Autumn is the one seasonal job worth putting in the calendar by name. Gutters and outlets on a shallow-pitched glazed roof block with leaf litter every year, and blocked drainage is behind a large share of atrium leaks that get reported as seal failures.
What has to be a specialist, and when
Anything requiring close access to the glass. Perimeter seal assessment, gasket condition, framework corrosion, fixing checks, weep hole clearance. None of that can be judged from below, and an inspection by someone who couldn’t reach the glass isn’t an inspection of the things that fail.
Twice a year suits most urban atria, annually works for clean inland sites with a sound history, and quarterly is proportionate near heavy traffic, industry or a flight path. Buildings with a record of leaks sit at the shorter end regardless of their age.
The access planning has documented duties attached under the Work at Height Regulations, which is a reason to have it done by people who do it routinely rather than by the general maintenance contractor with a ladder.
The record is the schedule’s real output
A schedule produces two things: work done, and information. The second one is worth more over time and gets looked after less.
What makes a record usable is comparability. Photographs from the same positions each visit. A defect register that carries items forward rather than starting fresh. A note of what was done about last time’s findings. Done that way, three years of reports show you which way the roof is going, which is the only basis on which anyone can plan a capital bid instead of reacting to a failure.
Done badly, you get a stack of reports that all say satisfactory and no idea whether the gasket at the north-west corner is worse than it was.
A schedule that fits on a page
Quarterly: a walk below with a look up, and a look at the walls and ceilings under the roof line. Autumn: clear gutters and outlets. Twice yearly, or at whatever interval the building justifies: specialist visit with close access, combining clean and inspection, with a comparable written record. After any storm: a look. Annually: review the defect register and decide what’s becoming capital rather than maintenance.
Contractors who hold the access and the glazing in one business, Sky Access among them, can usually deliver the combined visit under one scope and one permit.
If the schedule takes longer to read than that, the team won’t follow it, and an unfollowed schedule is worse than none because it reads like cover.