Leak Only On Wind-Driven And Heavy Rain Combined: Decision Tree

Why this matters

The hardest leak to find is the one that only shows up when wind and heavy rain hit at the same time. It passes every dry-day inspection and every gentle hose test, the customer thinks you missed something, and the next ordinary storm leaves no trace. The reason is mechanical: certain defects only admit water when wind pressure drives rain horizontally and uphill against a lap, or when a high enough rain rate overwhelms a drainage path. Diagnosing this requires reproducing pressure and volume, not just wetness, and knowing which details fail only under load. Get it right and you fix a leak nobody else could find.

Symptom presentation

The leak is rare and tied to specific storms: heavy rain with strong wind from a particular direction. Light rain, heavy-but-calm rain, and a normal hose test all stay dry. The interior stain often appears at a wall-roof transition, under a ridge or hip, at a rake edge, around a vent, or below the roof on the windward elevation. The customer can sometimes correlate the leak with storms from one compass direction, which is the single most useful clue you can get.

Quick checks

• Get the storm direction. Ask which way the wind was blowing during the storms that leaked. Wind-driven leaks are directional.
• Identify the windward features on that elevation: ridge vent, rake, headwall, sidewall, valley termination, vents, and the field headlap.
• Inspect the ridge vent for an external baffle and proper end treatment; baffleless or poorly ended ridge vents drive water in under wind.
• Check rake edges and drip edge laps on the windward side; wind lifts shingles and pushes water under rake-edge metal that laps the wrong way.
• Look at field shingle headlap and any high or backed-out nails; wind-driven rain can ride up under a marginal headlap that sheds fine in calm rain.

Isolation tree

Branch 1: Leak appears under or near the ridge. Wind blowing up the windward slope drives rain into a ridge vent that lacks an external wind baffle or has open, unshingled ends. Confirm by inspecting the ridge vent type and ends; a baffleless ridge vent on a high-wind elevation is a prime suspect. Heavy calm rain does not load it; wind does.

Branch 2: Leak at a sidewall or headwall, directional with wind at that wall. The step or head flashing passes a vertical hose test but wind pressure pushes water behind the vertical leg or under a missing kick-out. Confirm with a pressurized or wind-simulated spray (a calibrated spray plus a fan, or testing during an actual windward storm with an interior watch).

Branch 3: Leak at the rake or eave on the windward side. Wind lifts the first courses and pushes water under rake-edge or drip-edge metal that is lapped under the underlayment instead of over it, or shingles with inadequate starter adhesion. Confirm by lifting the windward starter and checking lap direction and seal.

Branch 4: Leak in the field with no penetration. Marginal headlap, high nails, or shingles not sealed (common on a cold-installed or recently replaced section) let wind-driven rain track uphill under the tabs. Confirm by checking nail height, headlap, and tab seal on the windward slope.

Branch 5: Low-slope section overwhelmed only by high rain rate plus wind ponding water against a curb or termination. Volume, not just pressure, is the driver. Confirm by checking drainage and termination height on the windward side.

A practical clue that narrows the search fast: defects that leak only under combined wind and volume almost always sit on the windward perimeter of the roof, not in the protected center, because that is where wind pressure and uphill water-driving are strongest. Map the windward edge, ridge, and wall lines on the storm-direction elevation first; the field interior is rarely the culprit for a load-dependent leak. This bias toward the windward perimeter cuts a large roof down to a few feet of detail worth testing.

Confirming diagnosis

A standard gravity hose test will not reproduce this leak; you must add pressure and volume. Use a calibrated spray method (AAMA 501.2 nozzle, or ASTM E1105 with an air-pressure differential for wall assemblies) directed at the windward detail, or stage a high-volume spray with a fan to simulate wind loading, always bottom-up and one detail at a time with an interior watch. The most reliable confirmation, when feasible, is to inspect during an actual windward storm. Reproduce both the direction and the intensity or the leak stays hidden.

Remediation

Fix the load-specific defect. Replace a baffleless ridge vent with an externally baffled type and shingle the ends correctly. Add the missing kick-out and correct step/head flashing at walls. Reset rake and drip-edge metal so it laps over the underlayment and seal the windward starter. Correct high nails and re-establish headlap and tab seal on the windward field. Raise terminations or correct drainage on an overwhelmed low-slope section. After repair, re-test with the pressurized spray method to confirm the detail now holds under load, not just under a calm trickle.

References

• AAMA 501.2, Quality Assurance and Diagnostic Water Leakage Field Check (nozzle spray method for fixed joints).
• ASTM E1105, Field Determination of Water Penetration by Uniform or Cyclic Static Air Pressure Difference (pressure-driven spray testing).
• 2021 International Residential Code R905.2.8 (asphalt shingle flashing) and R806 / ridge-vent provisions; manufacturer instructions for externally baffled ridge vents in high-wind exposure.
• NRCA Roofing Manual: Steep-slope Roof Systems, wind-driven-rain detailing for ridge vents, rake/drip edges, and wall flashings.