Dry in Flood Test but Leaks in Wind-Driven Rain: Misleading-Test Decision Tree

Why this matters

A roof that passes a flood test and then leaks in a wind-driven storm has not contradicted itself; the flood test simply does not test the failure mechanism. Flood testing proves the horizontal field and drains hold standing water under gravity. Wind-driven rain leaks come from vertical and near-vertical details (terminations, counterflashing, end laps, walls, penetrations) failing under lateral pressure that a flood test never applies. A tech who certifies a roof "watertight" on a passed flood test and is then surprised by a wind-driven leak has trusted the wrong test for the failure. Knowing the gap between the two lets you test the right detail and stop blaming the field for a flashing failure.

Symptom presentation

The roof was flood-tested and held water with no leak. Later, during a storm with strong wind, water enters, often at a wall, parapet, penetration, or termination rather than in the open field. The customer reports leaks only in windy storms, or only when wind comes from a particular direction, while quiet steady rain causes nothing. The interior stain typically appears near a vertical detail, not in the middle of a flat field.

Why the flood test misleads

A flood test impounds water on the horizontal surface to a shallow depth and checks for leaks under static, gravity-only conditions. It validates the membrane field, laps below the water line, and drains. It deliberately does not, and cannot, load the parts of the roof above the water line: base flashings, terminations, counterflashing, curb and penetration uprises, parapet caps, and wall transitions. Wind-driven rain attacks exactly those above-water details, pushing water laterally and upward against laps and terminations with pressure the flood test never creates. So a passed flood test is a true result for the field and an irrelevant result for the wind-driven detail.

Quick checks

• Map where the leak actually enters: a near-vertical detail (wall, curb, parapet, penetration, termination) versus the flat field. Vertical details point to wind-driven failure outside the flood test's scope.
• Correlate with wind: leaks tied to high wind, or to one wind direction, indicate lateral-pressure intrusion the flood test could not reproduce.
• Inspect terminations and counterflashing on the windward side: loose termination bars, failed reglet sealant, short or reverse laps, open end laps.
• Check penetration and curb base flashings for inadequate uprise height and unsealed tops where driven water gets behind.
• Note that the field passed; that result stands and steers you away from the membrane field toward the details.

Isolation tree

• Leak at a wall or parapet, worse with wind, field flood-passed: wind-driven intrusion at the base flashing, counterflashing, or cap. Inspect uprise height, reglet seal, and cap laps on the windward face.
• Leak at a curb or penetration in wind: driven water getting over or behind a short or unsealed uprise. Check flashing height and the sealed top termination.
• Leak only with wind from one direction: the windward detail on that exposure is the failure; the leeward identical detail may be fine, which is why steady rain misses it.
• Leak at an end lap or termination bar: lap shingled against the prevailing flow or a loose bar lets driven water track in.
• Field flood-passed, no vertical detail implicated, leak still tied to wind: suspect a deck-edge, gravel-stop, or coping detail; extend inspection to all above-water terminations.

Confirming diagnosis: test the real mechanism

Replace static flooding with a test that loads the windward detail under pressure.

• Pressurized spray-rack test per ASTM E1105: apply calibrated water to the suspect vertical detail while a static pressure difference simulates wind load; this reproduces wind-driven intrusion that a flood test cannot.
• Directed-stream check per AAMA 501.2 along terminations, counterflashing, and laps, understanding it is a directed-stream method best for locating open joints, not a substitute for pressurized testing.
• Windward-specific test: simulate the offending wind direction at the implicated detail; reproducing the leak only on that exposure confirms the directional, wind-driven mechanism.
• Re-flood is unnecessary; the field already passed. Spend the test budget on the above-water details where the leak actually is.

Remediation

• Repair the windward detail: raise short flashing uprises to adequate height, reseal or re-set counterflashing in the reglet, correct lap direction and overlap, tighten or re-bed termination bars, and seal open termination tops.
• Correct end laps and coping or cap joints on the windward face so driven water cannot track behind.
• Re-test with the pressurized or directional method on the repaired detail; a passed flood test is not proof of a wind-driven repair.
• Document for the customer that the field is sound (flood-passed) and the fix addressed the wind-driven detail, so the two results are not conflated again.

References

• ASTM D5957, Standard Guide for Flood Testing Horizontal Waterproofing Installations
• ASTM E1105, Standard Test Method for Field Determination of Water Penetration by Uniform or Cyclic Static Air Pressure Difference
• AAMA 501.2, Field Check of Metal Curtain Walls for Water Leakage
• ICC, International Building Code (IBC), Chapter 15 Roof Assemblies, Section 1503 Weather Protection and flashing provisions
• National Roofing Contractors Association (NRCA), The NRCA Roofing Manual: flashing details and leak investigation