My Commercial Roof Is Leaking, But Is It Actually the Roof?

Often it is not, and the reason is that almost all water intrusion happens at a tiny fraction of a building’s surface, at the junctions and terminations where one system hands off to another. In Oklahoma the usual culprits are a saturated block wall, open mortar joints, a parapet, a rooftop unit’s condensate line, or condensation that was never a leak at all. Arrowhead Roofing has worked commercial buildings in northeast Oklahoma since 1987, and the most useful thing we do on a persistent leak is often to tell an owner that the roof they were about to replace is not the problem.

Note: this is a general guide only. Every building is different and a diagnosis needs an inspection. Nothing here is a substitute for having someone look at the actual assembly.

Here is where water actually gets in, and what each route looks like from inside.

The Short Answer

The roofing industry’s own trade body publishes this, which is worth knowing before you read a roofer saying it.

The National Roofing Contractors Association’s magazine ran a piece on diagnosing persistent leaks that lists what actually turns up under the heading of a roof leak: mechanical issues related to heating and cooling equipment and commercial exhaust vents, exterior wall failures in veneers, doors and windows, drainage problems, and below grade waterproofing. Its central point is structural rather than accusatory: the enclosure has to work as a whole, and it is usually the transition points that fail.

Alongside that sits the line that gets quoted constantly in building enclosure work, and which is worth reproducing with its own hedge intact: it is said that ninety percent of all water intrusion problems occur within one percent of the total building exterior surface area. It is not the waterproofing materials that leak. It is the junctures and terminations between them.

A membrane is a large, continuous, factory made sheet. A parapet coping joint is a hand made detail. Only one of those is likely to be the problem, and it is not the one that gets replaced.

Water Travels Before It Shows Up

This is the single fact that misleads more building owners than any other, and it is why the stain is not a map.

Water that enters at one point migrates laterally through the assembly until it finds a route down. A published case study describes exactly that: water moving sideways through a roofing assembly until a penetration, in that instance a fall protection anchor, let it drop to the deck. Another documents water travelling through three fire walls along a channel at the top of a truss before overflowing onto a floor.

Two consequences follow. The stain on the ceiling tile is evidence that water arrived, not evidence of where it entered. And a water test has to be given time, because it can take hours for a leak to wet the whole path before anything appears inside.

The Block Wall Nobody Suspects

Concrete block is not waterproof, and it was never specified to be. That is a documented fact about the material rather than a defect in a particular building.

The masonry industry’s own technical guidance states that untreated masonry typically takes on water through absorption, adsorption and capillary forces, and that although block and mortar generally do not let water pass quickly, rain can pass through if driven by a significant force. It also says something that goes straight to the heart of this page: if rain enters a wall other than by way of the roof or at element interfaces, it can often be traced to the interface between the masonry unit and the mortar.

The governing product specification makes the same point from the other direction. It sets a maximum permitted water absorption for concrete masonry units, graded by density. A block that complies with the standard absorbs water. It is designed to, within a limit.

Single wythe block walls, which is what a great many light commercial buildings here are, have no drainage cavity behind them to catch what gets through. They depend on surface treatments, integral water repellents and drying. When those stop working, the wall becomes a reservoir, and a reservoir against the inside face of a building looks exactly like a roof leak.

What a Moisture Meter Reading Actually Means

This section exists because we use these instruments and because the way they are usually reported is misleading.

Hand held moisture meters are standardized against wood. The practice governing their field calibration and application covers the moisture content of solid wood and wood products. Point one at a concrete block wall and the percentage on the display is not a moisture content, because the scale it is reading against was never built for that material.

ASTM’s own guide for using an electronic meter on cementitious material makes this explicit in its title, which describes a comparative moisture condition, and in its text, which says it is not intended to provide quantitative results.

And there is no threshold to fail. Published work on moisture testing in walls states it directly: industry acceptance ranges for moisture content in walls are not available, and no guidance exists for the location and frequency of measurements. Acceptance criteria exist for floors. They do not exist for walls.

So what is the reading for? Comparison. Same instrument, same settings, same session, one location against another. On a job where the wall reads far higher than the assembly on the other side of it, that differential is real evidence and it is what a competent report should present. On one building we documented a reading of 61 percent on saturated block while the roof side of the assembly read low, which is what established that the water was coming through the wall. The number on its own means very little. The gap between two numbers means a great deal.

Parapets and Copings

The last few feet of a wall, and the place where a roof and a wall stop being two separate things.

A parapet takes weather on both faces, carries a coping with joints in it, sits at the transition where the roof membrane terminates, and is usually detailed by hand. Published enclosure work notes that when water shows up on a top floor it is often immediately attributed to the roofing system or the façade, when the parapet is the actual route. A documented courthouse case saw multiple roof replacements fail because the real problem was a saturated masonry parapet and missing through wall flashings.

Saturation there is not only a water problem. A parapet that cannot dry holds moisture through freeze and thaw cycles, and the expansion forces break the masonry apart from the inside. Spalled and cracked block at the top of a wall is usually a symptom rather than a cause.

When It Was Never a Leak

Condensation gets misdiagnosed as water intrusion routinely, and one specialist has written exactly that after seeing several cases of it.

Water vapour is deposited on a surface whenever that surface is colder than the dew point. Inside a roof assembly the same thing can happen out of sight, where vapour permeates the layers and condenses within the roof. Buildings that run high indoor humidity are the usual candidates, and a direct fired gas heater is worth knowing about because water vapour is one of its combustion products.

Two other things on the roof produce water without any rain at all. A rooftop unit’s condensate drain is one, and the mechanical codes require a secondary drain or auxiliary drain pan specifically where an overflow or a blockage would damage the building, which tells you how routine the problem is. Exhaust equipment is the other.

How you tell: a leak follows rain. Condensation follows temperature, humidity and occupancy, and condensate follows cooling load. Nobody has published that as a formal procedure and we are not going to pretend otherwise, but it is the first question worth asking, and it is free.

How a Real Investigation Runs

There is a published method for this, which surprises most owners, and it is the difference between diagnosis and guessing.

ASTM E2128 is the guide for evaluating water leakage of building walls, and it sets out a sequence: review the project documents, evaluate the design concept, determine the building’s service history, inspect, carry out investigative testing, analyze, and report. Each step is meant to inform the next. ASTM D7053 is the equivalent guide for determining and evaluating causes of water leakage on low slope roofs, and it is worth knowing that it expressly excludes condensation from its scope. The roof leak standard treats condensation as a different problem needing a different method.

The testing tools each answer one question and no more.

Spray and flood testing

Water applied to a defined area under controlled conditions, working from the bottom up so that a result identifies the lowest possible entry point. ASTM E1105 covers windows, storefronts and curtain walls with a calibrated spray rack and a pressure difference. ASTM C1601 is the masonry wall equivalent, applied to a wall surface at a specified flow rate and pressure.

Electronic leak detection

Low voltage and high voltage methods that find breaches in a membrane. ASTM D7877 covers them. And there is one limitation that matters more than all the others put together, published in the NRCA’s own magazine: electronic leak detection will not detect a leak that is not caused by a membrane breach. A clean result on this test is a result. It says the membrane is intact and the water is coming from somewhere else.

Moisture surveys

Infrared under ASTM C1153, nuclear under the ANSI standard, and electrical impedance under ASTM D7954. Each of the three locates areas of elevated moisture. None of them locates a leak, and all three require verification by cutting cores or extracting samples. Infrared in particular does not see moisture. It sees temperature differences, and moisture is inferred from them.

The Building Owner Who Was Certain

A worked example, because this is the shape most of these take.

An owner called about a leaking membrane roof and was confident about the diagnosis. Water was coming in, the roof was the thing above the water, and a replacement had been priced.

What the inspection found was that the membrane and its flashings were intact and still sealed. What it also found was saturated block on the exposed elevation and gaps in the mortar, with the meter reading far higher on the wall than anywhere on the roof side of the assembly. The water was entering the wall, moving through it, and appearing at the ceiling line, which is the point on the inside of a building where a wall problem is most easily mistaken for a roof problem.

The roof did not get replaced. That conversation is not a pleasant one to have with a salesperson’s quota in the room, and it is the reason to have the investigation done by somebody who is willing to have it.

Related Questions

What are the membranes and which one is on my building?

Three common ones, and they behave very differently here. TPO vs EPDM vs PVC: which commercial roof membrane is right? covers how to tell them apart and what each is for.

The building is a restaurant. Does that change anything?

Considerably, and the exhaust equipment is the reason. Why does a restaurant roof need PVC instead of TPO? covers what grease does to a membrane and what the code requires of the fan.

How old is the roof, and is that even the issue?

Age is one input and usually not the deciding one. How long does a roof last? covers the ranges and what actually ends a roof in this climate.

See Also

The two answers next to this one are what the membrane on most commercial buildings here actually is, and how the three membranes compare.