Repair, refurbish or replace a commercial roof — the honest decision

Three options, not two

The decision is rarely a straight repair-versus-replace binary. There are three real routes, and the gap between them is thousands of pounds and years of life.

1. Repair (and planned maintenance)

A localised repair addresses a specific, contained failure — a split at an upstand, a lifted seam, a blocked outlet, storm damage to a section — where the rest of the roof is sound. It is the right call when the deck, insulation and falls are in good order and only the surface has failed in one place. Done properly, a repair buys years for a small fraction of a re-roof.

Planned preventative maintenance (PPM) sits alongside it: an annual or twice-yearly inspection, gutter and outlet clearance, and detail repair. PPM protects a sound roof and its guarantee far more cheaply than reactive patching, and it catches a small failure before it becomes a wet-insulation problem. Indicative localised repair and overlay work runs around £25 to £70 per square metre — but repair is only value where the underlying roof is genuinely sound.

2. Refurbish (overlay, over-clad or coat)

Refurbishment recovers a sound existing roof without a full strip. On a flat roof that is an overlay — a new single-ply membrane or a cold-applied liquid system laid over the old one. On a profiled-metal roof it is an over-sheet or over-clad to a new insulated build-up, or a liquid coating with cut-edge corrosion treatment that buys 10 to 20 years. On a leaking industrial roof it is often a gutter lining, because the sheets are sound and the gutter has failed. All of these are cheaper and faster than a strip, avoid the disruption and waste of removing the old covering, and on an occupied building a cold-applied system removes naked-flame hot-works risk entirely.

The conditions are strict, though. Refurbishment only works where the substrate — deck, sheeting, insulation and falls — is sound and dry, and where the structure can carry any extra weight. It does nothing for wet insulation, and an overlay does not, on its own, correct a flat roof that ponds because it was never laid to fall. Where those conditions hold, refurbishment is often the smartest money on the table; where they do not, it is a patch with a longer name.

3. Replace (full re-roof or re-clad)

A full replacement removes the failed covering back to the deck and rebuilds the whole build-up — a new warm-deck flat roof with insulation and falls corrected, or a full re-clad to a new insulated composite or built-up system. It is the right call where the insulation is wet, the deck or sheeting is failing, a flat roof ponds because the falls were never designed, or a Part L thermal upgrade is due anyway. It is the most expensive route per square metre, but it is the only one that resets the roof to a full design life with a fresh manufacturer guarantee, and the point at which the U-value upgrade is designed in, typically to around 0.18 W/m²K.

The triggers — what points to which route

The build-up tells the real story, but the symptoms point the way before the survey confirms it.

Ponding. Standing water more than 48 hours after rain is the signature of a roof never laid to fall, or a deck that has deflected. Ponding accelerates membrane ageing and voids most guarantees. If the falls were never designed, no repair fixes it for long — this points to a strip-and-recover with tapered insulation to build the fall back in. BS 6229:2025 sets a minimum finished fall of 1:80, with the design fall set from structural analysis or a level survey, commonly 1:40 or steeper.

Recurring leaks. One leak in one place is a repair. The same roof leaking in different places, winter after winter, is a covering at the end of its life — patching it is spending on symptoms while the fault spreads. Recurring, migrating leaks point away from repair.

Life-expired covering. A membrane past its service life — 25 to 35 years for most single-ply and felt systems — will keep failing regardless of how well each patch is done. A building survey or dilapidations schedule flagging a life-expired covering is a replace signal, not a repair one.

Wet insulation. This is the decisive one, and it is invisible from the surface — core samples find it. Saturated insulation has lost its thermal value, it rots the deck from below, and it cannot be sealed over. Wet insulation rules out overlay and points to strip-and-recover.

Storm damage. Localised uplift or impact damage on an otherwise sound roof is a repair. Widespread uplift failure on a roof whose fixing pattern never suited its wind zone is a symptom of a covering that needs replacing to a proper BS EN 1991-1-4 uplift design.

A leak with sound sheets. On a large profiled-metal roof, an unexplained internal leak is most often not the sheets but a failed valley or box gutter, or cut-edge corrosion at the sheet laps. Both are far cheaper to put right than a re-clad — a gutter lining or cut-edge treatment often solves a leak that has defeated repeated sheet repairs. This is a refurbish signal, not a replace one, and it is exactly why a survey looks at the whole roof rather than just the reported drip.

A planned trigger. A lease event, a dilapidations schedule, a sale, a refit, or a decision to add solar are all planned triggers where the smart move is to survey now and act on evidence, rather than wait for the next ingress. If solar is on the horizon, the roof is surveyed and, if life-expired, re-roofed first — because no one wants to lift a new array to replace a failed membrane underneath it.

The whole-life sum that decides it

Once the survey has ruled options in or out, the choice between them is a whole-life calculation, not a headline price.

A life-expired roof patched reactively typically costs more over a ten-year horizon than a planned re-roof with a manufacturer guarantee measured in decades — before you count the business-interruption cost of a single major ingress. Reactive spend has no asset and no guarantee at the end of it; each patch is a call-out that treats the symptom while the fault spreads. A planned re-roof front-loads the cost but converts an escalating opex leak into a capital asset with a finite, transferable guarantee and a documented remaining life — worth real money at the next survey, sale or dilapidations event. See the full numbers on our cost guide.

Where the capital timing is the constraint, the works can be phased across financial years by roof area. A sound section waits; a failing section is done now. We will tell you honestly which sections can safely be deferred and which cannot.

Emergency repair versus planned works

There is one more distinction worth drawing, because it changes the economics: an emergency call-out and a planned programme are not the same purchase. When water is coming through over stock, IT or a trading floor, the first job is a temporary repair to stop the ingress — that is triage, and it is worth doing quickly. But an emergency patch is not a decision about the roof; it is a holding measure. The mistake is to let a run of emergency repairs become the strategy by default, because emergency work is the most expensive way to buy the least durable outcome. The right move after the first ingress is to survey calmly, establish what the roof actually needs, and make the repair-overlay-replace decision on evidence rather than under pressure. A planned re-roof is controlled, scheduled around your operation and priced properly; a string of emergencies is neither.

The tenant-and-landlord question

For a leaseholder, the decision often turns on the lease and the dilapidations schedule as much as the roof. Where the roof is demised to you or you carry a full repairing obligation, the liability is yours, and a planned re-roof is usually cheaper than the dilapidations claim at lease end. Where it sits with the landlord, an independent condition report and a costed remaining-life view let you serve or respond to a schedule with defensible figures rather than a guess. Either way, you get the evidence before anyone spends money.

A quick decision checklist

Before you commit either way, a survey should establish:

• The deck type and its residual structural capacity.
• Whether the insulation is dry or wet (core samples, not a surface look).
• Whether the roof is laid to fall, or ponds because it was not.
• The remaining service life of the existing covering.
• Whether the scope is notifiable and triggers a Part L U-value upgrade.
• On any pre-2000 building, whether an asbestos survey is needed before intrusive work.

Answer those honestly and the route is usually obvious: repair a sound roof with a contained failure, overlay a sound roof that needs a lifespan extension, and strip-and-recover a roof with wet insulation, failing falls or a life-expired covering.

How we approach it

We survey the roof first and give you repair, refurbishment and replacement side by side — with honest costs and remaining-life estimates for each — rather than defaulting to whichever is easiest to sell. That is the point of a survey-based quote over a patch-merchant's rate: you decide on evidence, and the number that goes to your board is one you can defend. If a repair or a gutter lining is the right answer, we will say so, even though it is the smaller job.