Roof repair vs replace: the threshold where repair stops paying

Roof replacement is one of the few five-figure household decisions with no natural trigger. Nothing fails on a date. The covering degrades gradually, the repairs get slightly more frequent, and most people decide either after a leak forces it or after somebody knocking on the door tells them to.

Neither is a good decision process. What follows is the threshold question stated properly: at what point does spending on this covering stop returning anything.

Age is the input everything else modifies

Manufacturer ratings on asphalt shingles describe performance under test conditions, not a promise about your roof. Actual service life is set by ventilation, installation quality, slope, tree cover and above all climate. A roof in a hail belt and the same roof in a mild coastal climate are not the same product ten years in.

One IBHS finding explains a surprise homeowners hit often. The insurance-industry research body reports that how well a shingle stays sealed is decisive for wind resistance, and that hail impact degrades that seal. The practical reading is that a roof which has taken hail is measurably less able to resist a later wind event, even where the hail did not obviously damage it. Ageing on a roof is cumulative and partly invisible.

So the honest use of age is as a modifier. Under about fifteen years, assume repair unless something specific says otherwise. Past twenty, assume the covering is in its final phase and read every new problem in that light.

Four signals that the covering is finishing

Row three deserves the weight. A single leak tells you about one detail that failed. Leaks appearing at unrelated locations within a couple of years tell you the covering as a system has stopped performing, and no amount of patching addresses that.

The arithmetic that settles it

Round illustrative numbers chosen to show the method, not market averages. Take the quoted repair, divide it by the years of service you would honestly expect to buy from it, and compare that with the replacement quote divided by the expected life of a new roof.

A $1,200 repair buying three more years on an eighteen-year-old roof costs $400 a year. A $14,000 replacement over an expected twenty-five years costs $560 a year. On those figures the repair still wins, and the correct answer is to repair and plan.

Change one input and it inverts. If the same repair buys only one more year because the covering is failing generally, it costs $1,200 a year and the replacement is plainly cheaper. The number that decides this is not the price of the repair. It is how long you actually believe it will hold, and a contractor who will not put that estimate in writing is telling you something.

Two things sit outside this arithmetic. Deck damage found during tear-off is not optional and not always visible in advance. And a covering at the end of its life is worth replacing on your timetable, in a slow season, rather than during the week after a storm when every crew in the region is booked and pricing accordingly. What actually moves the replacement figure is set out in our guide to replacement cost drivers.

Three things that are not replacement signals

Being fair to the roof matters as much as being honest about it, because several conditions that get sold as terminal are not.

Algae streaking. The dark vertical staining on north-facing slopes is a cyanobacterium feeding on limestone filler in the shingle. It is cosmetic, it does not shorten service life meaningfully, and it is routinely presented at the door as evidence of a failing roof.

A few missing shingles after a storm. That is a repair, and often a claimable one. It becomes a replacement argument only where matching material is genuinely unavailable, which is a settlement question rather than a condition question.

Moss on a shaded slope. Worth removing carefully because it holds moisture against the surface, but growth alone does not mean the covering underneath has failed. Pressure washing an asphalt roof to remove it, on the other hand, does real damage by stripping granules, and is one of the more common ways a serviceable roof gets shortened.

None of this means a roof showing these is fine. It means these three should not be the reason a replacement gets signed, and if they are the headline in somebody’s pitch, that is worth noticing.

Before you accept that it is time

1. Get the age from the permit record, not from a guess or from whoever is quoting. Most municipalities hold the re-roof permit.
2. Get two or three assessments from firms that will also quote repairs. A business that only sells replacements will find replacements.
3. Ask each to state, in writing, the remaining life they expect from a repair. That single number drives the arithmetic above.
4. Establish whether any of it is storm-related before committing to pay for all of it. Our note on storm damage against wear covers how that call is made, and age-related deterioration is generally excluded.
5. Then decide how to pay for it, which is a separate question worked through in our piece on claim against cash against borrowing.

Questions readers actually ask

Does a 30-year shingle last 30 years?

Treat the rating as a product description under test conditions rather than a service-life promise. Real duration depends on ventilation, installation, slope, shade and climate, and a roof that has taken hail ages differently again. Read the number as a relative quality indicator between products, not as a date on your calendar.

Can I put new shingles over the old ones?

Sometimes, and it is cheaper because it removes the tear-off and disposal cost. It also hides the decking, so nothing underneath gets inspected, and it adds weight. Local code limits how many layers are permitted. The saving is real and so is the trade, which is worth naming rather than letting it appear as a mysteriously low bid.

How many leaks are too many?

The count matters less than the locations. Repeated leaks at the same detail suggest that detail was repaired badly. Leaks appearing at unrelated points within a couple of years suggest the covering has stopped performing as a system, and that is the point at which further patching stops returning anything.

Is grit in the gutters a problem on its own?

A new roof sheds loose granules for a while, so on a young roof it is normal. On an older one, steady accumulation means the mineral layer protecting the asphalt is thinning, and IBHS treats granule loss as a genuine damage mode. Combined with age it is a late-stage signal rather than a maintenance item.

Written by

Keith Guirao

Founder & Editor, ConsumersWeek

18+ years in consumer marketing and lead generation across insurance, personal finance, and home services. ConsumersWeek explains how these products are priced and sold so you can evaluate them with the same information the industry has.

About the editorLinkedIn