We have walked roof inspections after every named storm to hit our service area for the last decade. The pattern is consistent enough that the lessons are worth writing down — even with the obvious caveat that no two storms are identical and no individual roof's outcome can be predicted from generalities.
What the data tells us
Across the post-storm inspections we have conducted, the failure rate by roof type runs approximately:
- Older asphalt shingle, pre-2007 attachment: 35-55% of roofs damaged enough to require partial or full re-roof
- Newer asphalt shingle (2007-2020), code-compliant attachment: 12-22% damage rate
- Concrete tile, RAS 118 (mortar): 25-40% damage rate
- Concrete tile, RAS 120/127: 8-15% damage rate
- Clay tile, RAS 118: 18-30% damage rate (similar to concrete but with smaller piece loss)
- Clay tile, RAS 120/127: 5-12% damage rate
- Standing-seam metal, mechanical-lock: 3-8% damage rate
- Standing-seam metal, snap-lock: 10-20% damage rate
- Natural slate: 5-10% damage rate (and the damage is almost always to flashings, not to the slate itself)
These are storm-specific approximations, not warranties. But the differential is consistent enough to inform specification decisions.
What we have learned
Attachment matters more than material
A correctly attached RAS 127 concrete tile roof outperforms a poorly attached RAS 118 clay tile roof. The material is the smaller factor; the attachment is the larger.
Edge zones fail first
Failure almost always starts at the perimeter — eaves, rakes, hip lines. The field of the roof, once the perimeter is intact, generally holds. The first three courses of tile or shingle from the edge are the leverage points for upgrade.
Secondary water resistance saves homes
The single largest difference between "damaged roof" and "destroyed home" is whether secondary water resistance was in place beneath the field material. A wind-stripped roof with full self-adhered underlayment is repairable. A wind-stripped roof with felt underlayment becomes a destroyed home.
Metal mechanical-lock is the most reliable in major events
In our post-storm data, mechanical-lock standing-seam metal has the lowest damage rate. The pieces are interlocked mechanically; the failure mode is structural (panel pulls off in a section) rather than individual (tiles peel off one by one). When metal fails, it is dramatic. When it does not fail, it is essentially intact.
Slate is the least visible failure mode
Slate roofs in major events typically lose flashings before they lose stone. The slate itself is anchored in a way that distributes wind load across the field. Flashing failure is repairable; the field survives.
What this means for new specifications
For new construction or re-roof in Florida's wind zones, we specify:
- Self-adhered underlayment over the entire deck, always
- Attachment one step tighter than code-prescribed for the zone
- Premium fasteners (stainless on the coast, ring-shank inland)
- Mechanical-lock seaming on any metal roof
- Hand-set flashings, not extrusion-cut
The cost premium over budget specifications is meaningful but not transformative. The outcome differential after a major event is substantial.
A caveat
Every storm is its own event. The 35-year-old asphalt roof that surprised us in 2017 by holding up to a Category 4 was not predicted by our generalities. The newer code-compliant clay roof that surprised us in 2022 by losing 30% of its field was not predicted either.
What we can predict is the average. And the average is meaningfully better with the upgraded specifications we have learned to write.
Filed under