A roof maintenance plan is a document you own that says which roofs you have, what happens to each of them, when, who does it, and where the record goes afterwards. It is not the contractor’s proposal. The proposal is one section of it, and it gets replaced when the contractor does.
Most buildings do not have one. They have a contractor, an arrangement, and a shared assumption that somebody is watching the roof. That works until the person holding it leaves, until the contract ends, or until an insurer asks a question that only a document can answer.
Start with the inventory
You cannot schedule what you have not listed. Before anything else, write one page per roof covering the facts that decide everything downstream.
| Field | Why it is there |
|---|---|
| Area and number of sections | Sections of different ages are separate roofs with separate schedules |
| System and approximate age | What fails on it, and how much life it is expected to have left |
| Warranty status and conditions | Whether inspection frequency and approved contractors are already decided for you |
| Access | Hatch, ladder, lift, permit, tenant route. This drives cost on every single visit |
| Drainage type | Internal outlets, scuppers, gutters, and whether overflows exist |
| What is on it | Units, ducts, solar, screens, walkways, anything bolted through |
| What is under it | The consequence of a leak, which is what sets priority |
| Known history | Leaks, repairs, surveys, and who holds the paperwork |
Where you do not know, write unknown rather than a guess, and make establishing it a task with a date on it.
Set the frequency by the roof
A schedule that gives every roof two visits a year is a filing convention, not a plan. Frequency follows risk, and risk on a roof comes from four things: its age, how many people go on it, how much equipment sits on it, and what it has already done to you.
A young roof with nothing on it and a locked hatch needs less attention than a roof of the same age carrying kitchen extract, a solar array and a service contractor every month. Most damage on a busy roof is caused by people, and people arrive all year rather than seasonally.
Write the triggers, not only the dates
Half the visits that matter are not on the calendar. The plan should name the events that cause a visit regardless of when the last one was.
- Severe weather: wind, hail, heavy or prolonged rain, a significant snow load
- Any trade working on the roof, before they leave and after
- A new stain, drip or damp patch reported from inside
- A tenant fit out, a new rooftop unit, or an array going up
- The roof reaching an age or a warranty date that calls for a formal survey
Name who can raise a trigger and who acts on it. A trigger nobody owns is a paragraph.
Say who does what
Three parties do the work and the plan should be explicit about which is which.
The owner or facilities team controls access, keeps the log of who went up, receives and files the reports, raises triggers, and holds the budget. None of that is a contractor’s job and none of it happens by default.
The maintenance contractor walks the roof, clears the drainage, carries out the small corrective work agreed in their scope, reports with photographs and flags what needs a separate quote.
Everyone else on the roof works to a written condition: designated routes, no fixings through the membrane, damage reported the same day. Put that in writing to the mechanical, telecoms and solar contractors, because it changes behavior.
Scaling it to a portfolio
A portfolio plan is the same document repeated, with three additions that only matter at scale.
One format. The same page layout, the same condition language and the same rating scale for every roof, so two buildings can be compared rather than described.
One register. A single list of every roof with its last visit, next visit, condition rating and open findings. This is the page that goes to whoever approves the budget.
One order. A ranking of the roofs by consequence and condition, with the reasoning written next to each. When money is short, the ranking decides what gets deferred, and a ranking with reasons attached survives that conversation.
Keep it short enough to be used
One page per roof, one register, one review date. A plan that runs to a binder gets filed, and a filed plan is the same as no plan on the morning somebody needs to know whether the roof over the warehouse has ever been surveyed.
Ready to put it on a schedule? See how a maintenance program runs. Walking the roofs yourself first? Start with the roof maintenance checklist.
Common questions
Should we write the plan or should the contractor?
You write the plan. A contractor writes a scope of work, which is what they will do for you. The plan is the wider document that says what has to happen to the roof, whoever does it.
- The plan survives a change of contractor, a scope of work does not
- It holds the things no contractor can own: access control, records, budget
- A contractor should help you write the technical parts and often will
- Their scope then sits inside your plan as one section of it
The practical test is what happens when the provider changes. If the new contractor has to rebuild the picture from scratch, what you had was a contract rather than a plan, and the building has just lost its history at the worst possible moment.
What do we put in the plan for a roof nobody knows anything about?
Write unknown, and make finding out the first scheduled task. A plan with honest gaps is useful. A plan with guesses in it is worse than no plan.
- Record what can be seen: area, access, drainage type, visible system
- Record what is missing: age, warranty, as built drawings, repair history
- Book a survey to establish the rest, and date the entry when it arrives
- Treat an unknown roof as higher priority until it has been looked at
Most portfolios have at least one of these, usually acquired with a building. The gap matters most when something fails, because the first questions asked by a warranty provider or an insurer are exactly the ones an unknown roof cannot answer.
How do we decide which roofs come first in a portfolio?
By consequence and by evidence, not by age. Rank each roof on what happens if it leaks and on what its condition record says, then let the two together set the order.
- Consequence: what sits under it, and what stopping that would cost
- Condition: leak history, repeat repairs in the same area, survey findings
- Unknown roofs rank high until somebody has established which column they belong in
- A sound roof over a server room can outrank a tired one over a store
Write the ranking down with the reasoning beside it. The ranking is what you defend when a budget is cut, and a list of roofs in order with no reasons attached is the first thing to be overruled by whoever shouts loudest.
How often should the plan itself be reviewed?
Once a year as a fixed date, and immediately after anything that changes the facts in it.
- A storm, a major repair, a replacement or a new roof at handover
- A change of tenant, of use, or of what sits under the roof
- New rooftop equipment, a solar array or a screen going up
- A change of contractor, of insurer or of warranty status
The annual review is the moment to ask whether the visit frequency is still right, because roofs age into needing more attention rather than less. Put the review date in the plan itself so it does not depend on anyone remembering.
How is a maintenance plan different from a capital plan?
A maintenance plan covers the work that keeps a roof alive this year. A capital plan covers replacing it. The maintenance plan feeds the capital plan, which is most of its value to a finance team.
- Maintenance is recurring, predictable and small
- Replacement is a single large sum that has to be known about in advance
- Condition records are what turn a guess at timing into a forecast
- A roof with a record can be scheduled, one without it gets discovered
Keep them as two documents that reference each other rather than one that tries to do both. Ask your maintenance provider to state, after each visit, whether anything they saw moves the expected replacement window earlier or later, and carry that line into the capital plan.