Wind-Driven Rain: Why Coastal Roofs Leak at the Edges

Picture a shingle course. The exposed portion sheds water downward. Underneath its lower edge is a gap — small, but continuous along the entire course — where the shingle above overlaps the one below.

In ordinary rain that gap is irrelevant, because water has no reason to travel upward into it. In a coastal storm, wind pressure against the roof surface provides exactly that reason. Water is pushed laterally and upward, into the gap, past the sealed edge, and onto whatever is underneath.

What is underneath is the underlayment, and this is the entire point. In a mild climate the underlayment is a secondary layer that mostly matters during construction. On a Mount Pleasant roof in September, the underlayment is doing real work several times a year, and the difference between a roof that leaks in a storm and one that does not is frequently a decision made about a layer nobody will ever see again.

The same pressure differential drives water sideways at valleys, behind wall flashing, and around pipe penetrations. Every one of those is a place where the shedding logic has been interrupted by geometry, and every one needs a detail that does not rely on gravity alone.

There are effectively three products in this conversation and they are not interchangeable.

Asphalt-saturated felt is the traditional option. It works, it is inexpensive, and it has two weaknesses that matter here. It wrinkles when it gets damp, which telegraphs through the finished shingle surface, and it tears readily around fasteners under wind load.

Synthetic underlayment is a woven polymer sheet. It is lighter, dramatically stronger in tear resistance, does not wrinkle the same way, and holds up far better if it sits exposed for a day or two between tear-off and dry-in. Given the reliability with which an afternoon thunderstorm interrupts a roofing job in this town, that exposure resilience is a practical benefit rather than a specification detail.

The component that matters most is neither of those. It is the self-adhered membrane, sold almost everywhere as ice-and-water shield, which is an unhelpful name in South Carolina. Its distinguishing property is that it is adhesive on the back and it self-seals around any fastener driven through it. That is the whole trick. A nail through felt is a hole. A nail through a self-adhered membrane is a sealed penetration.

In a wind-driven rain event, water that gets under the shingles arrives at a surface perforated by thousands of fasteners. Whether that surface seals around them is the difference between a wet underlayment and a wet ceiling.

This is the single most useful question to ask about any coastal roofing proposal, because the answers vary enormously and the price difference is real.

The minimum common specification is valleys only. Valleys concentrate the flow of two roof planes into one channel, so it is the obvious place to start and the cheapest way to say the word included.

A better specification adds the eaves. The eave is where wind-driven rain is pushed hardest against the direction of the overlap, and it is also where a backed-up gutter can hold water against the roof edge. Eave protection is doing double duty.

Better still adds the rakes — the sloped edges — which on a gable end take the full lateral force of wind hitting the side of the house, and every penetration: pipe boots, vents, the perimeter of any skylight, and both sides of every wall transition.

The most robust specification, and the one that FORTIFIED designations are built around, is a fully sealed roof deck, where the entire deck is covered in a self-adhered membrane or the deck seams are taped before underlayment. This is a genuinely different level of protection, because it stops treating water intrusion as an edge problem and treats the deck as the actual water barrier.

None of those four are wrong. They are four different products at four different prices. What is wrong is not knowing which one you bought.

Drip edge is the L-shaped metal at the eaves and rakes. It supports the shingle at the roof edge, keeps runoff from curling back under the edge by surface tension, and directs water into the gutter rather than behind it. The residential code requires it on asphalt shingle roofs.

It is one of the least expensive components on the entire roof, and it is occasionally reused on a re-roof because the existing one is still physically present. That is a false economy in coastal exposure. Old drip edge has usually taken corrosion at the fastener points, has often been bent during the tear-off, and no longer sits tight to the deck along its length. A drip edge that does not sit tight is an open invitation at exactly the point where wind pressure is highest.

There is also a sequencing detail that separates careful work from fast work. At the eave, drip edge goes under the underlayment. At the rake, it goes over. That order exists so that water always lands on top of the next layer down rather than behind it. It is a small thing, it takes no extra time, and getting it backwards produces a roof that leaks at the edges in exactly the storms it was supposed to handle.

Ask whether new drip edge is included, and if you want to know how carefully a crew works, ask which way it is lapped at the eave versus the rake. The answer will tell you a great deal.

Wherever a roof plane runs into a vertical wall — a dormer cheek, the side of a chimney, the wall of a bump-out or an addition — the shedding logic stops. Something has to carry water out of that junction and back onto the roof surface.

The correct detail is step flashing: individual L-shaped pieces, one per shingle course, each woven into the course and lapped over the one below, running up the wall behind the siding or counterflashing. Because each piece is interleaved with the shingles, step flashing comes off with the roof and must go back new.

The shortcut is to leave the old step flashing in place and run a heavy bead of sealant along the wall-to-roof joint instead. It looks fine. It sheds water on the day it is done. And it is a maintenance item with a service life of a few years, installed on a roof intended to last decades. A wall transition depending on caulk is on a countdown from the moment the crew leaves.

This is one of the highest-value questions you can ask, because it is invisible afterward and it is a common place to save half a day of labour. Ask directly: is the wall flashing being replaced, or sealed. There is no ambiguous third answer.

The related detail is counterflashing at masonry. Where a chimney meets the roof, the step flashing needs a second piece let into a groove cut in the mortar joint, overlapping it. Surface-mounted counterflashing stuck to the face of the brick with sealant is the same countdown in a different location.

Every hole in a roof is a managed failure. Plumbing vents, exhaust fans, electrical masts, skylights — each one interrupts the shedding surface and relies on a detail to compensate.

The plumbing vent boot is the most reliable early failure on any asphalt roof. The collar that seals against the pipe is typically rubber, and rubber under continuous ultraviolet exposure hardens, shrinks and splits. It routinely fails well before the shingles do — a roof can be barely into its second decade with a boot that is already cracked open.

Because of that, replacing every boot during a re-roof is close to free relative to the cost of the job, and leaving old ones is a genuinely bad decision. Ask whether new boots are included. If the answer is that the existing ones look serviceable, you are being handed a five-year clock on a thirty-year roof, and the leak will appear in a ceiling rather than on the roof.

Where longevity matters, boots with a lead or a metal collar rather than a moulded rubber gasket last substantially longer, because the failure mode is mechanical rather than ultraviolet. On a coastal house with an expensive interior finish below, that upgrade is a rational purchase.

Nailing is not a component you can point at, which is why it is almost never discussed, and it is one of the strongest determinants of whether a roof survives a wind event.

Three things matter. The number of fasteners per shingle, whether they land in the manufacturer's designated nailing zone, and whether they are driven to the correct depth. High-wind applications generally call for more fasteners per shingle than standard, and manufacturers publish the requirement.

Placement is where speed shows up. A nail driven high, above the nailing strip, secures only one course instead of catching the head of the shingle below, which means the fastener is doing roughly half the job it was meant to do. From the ground, a roof nailed high looks exactly like a roof nailed correctly, until a sustained wind finds it.

Depth matters in both directions. Overdriven nails cut through the mat and lose their grip. Underdriven nails leave a proud head that lifts the course above it and interferes with sealing.

You cannot inspect this after the fact without lifting shingles. What you can do is ask what fastening pattern is being used and whether it meets the manufacturer's high-wind requirement, and take the fluency of the answer as a signal.

The components in this chapter are all invisible by the second day, which makes the written scope the only durable record of what you bought.

Specify underlayment type by product category, not by the word quality. Specify exactly where self-adhered membrane is applied — valleys, eaves, rakes, penetrations, or full deck. Specify new drip edge at eaves and rakes. Specify that all wall and chimney flashing is replaced rather than sealed, and that counterflashing at masonry is let into the mortar. Specify new pipe boots and their type. Specify the fastening pattern.

That list fits on half a page and it converts most of the ambiguity in a roofing proposal into something you can hold someone to. It also, in practice, sorts contractors quickly. The ones who work this way will fill it in without hesitation, because they were doing all of it anyway and are pleased to be asked.