Repair or replace is not a judgment call. It is a conclusion drawn from four facts, and the reason it feels like a judgment call is that most owners are asked to decide before anybody has established them.
The four facts are how much of the roof is wet, what condition the membrane is in across the field rather than at the leak, whether the deck is sound, and what the repair history has been doing over the last few years. Get those, and the answer is usually obvious. Skip them, and you are choosing between two numbers from two contractors who are describing different roofs.
Fact one: how much of the roof is wet
Membrane is a surface. Insulation is the part that holds water, and once water has moved sideways through it the wet area is far larger than the opening that let it in. No surface repair dries it out.
This is established by a moisture survey, with the wet areas marked on a roof plan and the method stated. Ask how the wet areas were identified and how they were confirmed, because a scan alone produces anomalies rather than findings. Then ask for the removal of that material to be priced as its own line. It is the single figure that moves the decision most.
Fact two: the membrane across the field
A roof can leak in one place and be sound everywhere else. It can also be leaking in one place and finished everywhere else. The difference is found away from the leak.
The evidence is physical and it is gathered by hand: seams probed along their length rather than looked at, the membrane checked for brittleness and crazing, the perimeter checked for shrinkage pulling the sheet away from walls and curbs. Shrinkage across the field is close to a hard stop, because there is no repair that gives a membrane its dimensions back.
Fact three: the deck
The deck decides whether you are buying a roof or a construction project. Soft spots underfoot, corrosion where steel has stayed wet, fasteners that no longer hold, visible deflection: any of these turns a membrane job into structural work, and it is much cheaper to know before the tender than after the strip.
Deck condition is confirmed by core cuts, located on the same plan as the moisture survey and closed the same day. It is the one fact that cannot be established without opening the roof somewhere.
Fact four: the trend in what you spend
Pull the repair invoices for the last few years and mark each one on the roof plan. Two patterns matter.
Repeat repairs in the same area are the strongest single signal that the problem is the roof rather than the event. A roof that has been fixed three times in one bay is telling you the fix was never the issue.
Rising annual spend across the whole roof is the other. When the yearly repair figure starts to look like a meaningful share of replacement cost and the line is going up, you are paying for a replacement in instalments and not receiving one.
What is not evidence
- A free inspection from a firm that sells replacements. Get the survey from someone who is not quoting the work.
- An aerial photograph or a drone pass. Useful for layout and for standing water. It cannot tell you whether the insulation is wet.
- The age of the roof. A prompt to look, never a conclusion.
- One bad winter. A storm creates damage. It does not, by itself, end a roof.
When the facts point both ways
Two situations come up often enough to name.
The membrane is tired but the insulation is dry. That is the case restoration exists for, and it is usually the cheapest of the three options on a roof still in that condition.
The building has several roof sections of different ages. Then the question is not repair or replace at all. It is which section, and in what order, which turns an unaffordable project into a phased one. Each section keeps its own record and its own schedule afterwards.
Before you approve either
Ask for the wet areas on a plan, the cores photographed and located, the deck condition stated plainly, the removal of wet material priced separately, and the scope written to say what happens if more wet material is found once the roof is open. The commonest dispute on any roof job is not the quality of the work. It is a quote written for a dry roof and an invoice written for a wet one.
Still deciding what the survey should cover? Read roof leak detection methods. Repairing for now and planning ahead? Start a roof maintenance plan.
Common questions
How much of a roof has to be wet before replacement is the answer?
There is no single figure, and anyone who gives you one without seeing the survey is guessing. What decides it is how much wet material has to come out, and whether taking it out leaves a roof or a patchwork.
- Isolated wet areas are cut out and replaced during a repair
- Wet areas scattered across the field mean the assembly is compromised
- The cost of cut and replace work climbs quickly with the number of areas
- At some point it exceeds the cost of a new roof laid in one pass
Ask the surveyor to mark the wet areas on a plan and price the removal separately from everything else. Once that line is visible, the decision usually makes itself, and it makes itself on evidence rather than on opinion.
Does the age of a roof decide whether to replace it?
No. Age sets an expectation and nothing more. Roofs of the same age and system routinely differ by years of remaining life, depending on what has been done to them and what has been left alone.
- A maintained roof with dry insulation can outlive its expected span
- A neglected one can be finished well inside it
- Age matters most as a prompt to survey, not as a verdict
- It also matters for warranty status, which is a date rather than a condition
Use the age to decide when to start looking rather than what to conclude. A roof past the midpoint of its expected life should be surveyed for moisture while it is still behaving, because that is the last point at which every option is still open.
Can one section be replaced and the rest kept?
Often, yes, and on a building with roofs of different ages it is usually the right answer. What matters is the junction between the new work and the old.
- A tie in between two systems has to be designed, not improvised
- New and old sections will move differently and age differently
- Each section keeps its own record, its own warranty and its own schedule
- Doing sections in sequence spreads the cost across budget years
Ask how the tie in will be made, what happens to the drainage across the boundary, and whether the manufacturer will warrant work that abuts an existing roof. Phased replacement fails at the joins far more often than it fails in the middle of a section.
We were told the roof is beyond repair. How do we check that?
Ask for the evidence behind the statement, in writing. Beyond repair is a conclusion, and it should rest on facts that somebody established rather than on a walk around the parapet.
- A moisture survey with the wet areas marked on a plan and the method stated
- Photographs of the membrane condition across the field, not only at the leak
- Core cuts showing the layers, located on the same plan
- The deck condition, and how it was checked
If those do not exist, what you have is an opinion from somebody who sells replacements. Get a survey from a party who is not quoting the work, and compare the two. A contractor confident in the conclusion will not object to the roof being looked at by somebody else.
Is it worth replacing a roof that is not leaking?
Sometimes, and it is usually the cheapest version of the job. A roof replaced before it fails is a planned project. The same roof replaced after it fails carries wet insulation, damaged contents and a deadline.
- Planned work can be tendered, scheduled and phased across budget years
- It happens in good weather with the building protected
- A dry deck and dry insulation keep the scope predictable
- Emergency replacement removes every one of those advantages
The case is easiest to make when the condition record shows the trend rather than a single opinion. That is what a maintenance history is for, and it is why the forecast matters more to a finance team than this year's report does.