Field Water Testing: How to Prove Which Detail Is Actually Leaking

A ground-floor suite gets water at the window head during every heavy rain. The strata calls the window supplier, who looks at it and says the sealant has failed. The sealant contractor comes out and says the sealant is fine, the problem is the flashing above. The roofer says the roof is not involved at all. Everyone is confident, nobody is paying, and the owner in that suite has had a bucket on the sill for two winters.

This is the ordinary state of a leak investigation when nobody has done a controlled test. The building leaks, that much is agreed. Which detail is letting the water in remains an opinion held with varying degrees of conviction by people with an interest in the answer.

Field water testing is how that argument gets settled.

The principle is isolation

Rain wets the whole building at once. When water shows up inside, it could have entered at the window perimeter, the roof-to-wall junction, a failed sealant joint two floors up, a wall penetration, or a crack in the cladding. Water travels sideways and downward inside an assembly before it appears, which is why the visible wet spot is so often nowhere near the entry point. That behaviour is covered in the post on why leaks show far from the source.

A field water test reverses the problem. Instead of waiting for weather and guessing backwards from the wet spot, water is applied deliberately to one detail while everything above and around it stays dry. Someone watches the interior. If water appears, that detail leaks. If it does not, the water is coming from somewhere else and the next detail up gets tested.

The result is a fact rather than a theory, and it is documented, repeatable, and usable in a warranty claim.

ASTM E1105: the one with the pressure chamber

This is the more rigorous of the two common methods, and the one that most closely reproduces how leaks actually happen.

A chamber is sealed to the interior face of the window or wall area being tested. A calibrated spray rack wets the exterior at a defined rate, and a blower holds a pressure difference across the assembly for the duration of the spray, typically fifteen minutes of continuous water.

That pressure difference is the whole point. Water sitting on a joint does not go anywhere on its own; it needs a driving force. In a real storm, that force is wind pushing rain against the building, and it is what turns a marginal joint into a leaking one. A test that just wets a window without any pressure difference will pass assemblies that fail reliably in a November southeaster. The mechanism is the same one described in the guide on wind-driven rain.

The pressure used should be stated in the test report, because the result only means something at that pressure. A window that holds at a low test pressure and leaks at a higher one has told you something useful about where its limit sits.

AAMA 501.2: the nozzle test

This one is simpler and used for a different purpose. A calibrated nozzle directs a concentrated stream of water at a fixed joint from a set distance, moving along the joint at a controlled speed of about five feet per minute.

There is no chamber and no pressure difference. It tests the joint on its own terms: whether that sealant line, that gasket, or that glazing seal holds water directed straight at it.

It applies to fixed, non-operable glazing and to sealant joints. It is not appropriate for operable windows or doors, where the weatherstripping and drainage design are meant to handle water differently, and where a concentrated stream will produce a failure that means nothing about real performance.

It is faster and cheaper than E1105, and it is the right tool when the question is narrow: does this specific sealant joint leak. The sealant testing methods in the post on field pull testing of sealant adhesion answer a related but separate question, which is whether the sealant is still bonded rather than whether the joint keeps water out.

Preparation is most of the work

The spray itself takes fifteen minutes. The day it consumes is spent on everything around it.

Somebody has to be inside the assembly, watching, for the entire test. That usually means interior finishes come off first. If drywall, insulation, and window returns are left in place, water entering the cavity gets absorbed, travels, and surfaces somewhere else two days later. The test runs, nothing appears, and the assembly is declared sound when in fact it filled the wall cavity with water.

On a serious investigation, removing finishes ahead of the test is what makes the result mean anything.

Access is the other cost. On a high-rise, the spray rack and the technician operating it have to get to the exterior face at height. That is a rope access or swing stage setup with its own planning, and the choice between them is set out in the comparison of rope access versus swing stage. The access arrangement often costs more than the testing itself, which is why a sensible programme groups several test locations into one mobilization.

The owner also needs to agree in advance how much water is acceptable inside the test area. A test on an assembly that is known to leak badly will put water into the building on purpose. Containment, protection of finishes, and a plan for drying come before the hose turns on.

Choosing what to test

A test result applies to what was tested, and nothing more.

If the question is whether one specific window in one specific suite leaks, one test answers it. If the question is whether the whole elevation of windows was installed correctly, one passing test proves little. Windows on a tower are installed over weeks by different crews in different weather, and the variation between them is exactly what a sampling programme is meant to capture.

A representative sample spreads across floors, orientations, and where records allow, different installation crews or dates. The southwest elevation takes the worst of the driving rain in Metro Vancouver, so it belongs in any sample. So does at least one location on a sheltered face, because a leak there points at workmanship rather than exposure.

Where a test programme is being run as part of a larger investigation, it sits alongside other methods rather than replacing them. Thermal imaging finds candidate areas quickly and cheaply, as described in the post on thermal imaging envelope surveys, and water testing then confirms or rules out the specific details that scan flagged. Using the cheap method to narrow and the expensive method to confirm keeps the budget sensible.

Reading the result honestly

Two failure modes come up repeatedly in how results get used.

The first is over-extension. A window passes, and the report is waved around as proof the windows are fine. It proves one window did not leak at one pressure on one day. Anyone extending it to the building is going beyond what the test supports.

The second is the false pass caused by inadequate preparation. Water goes into the wall, is absorbed by insulation, and nothing shows at the observation point within the test window. The assembly gets a pass it does not deserve. This is why removing interior finishes matters, and why a test that cannot be prepared properly is sometimes worse than no test, because it produces a confident wrong answer.

A genuine negative result is useful. Knowing the windows hold means the investigation moves upward to the roof-to-wall junction, or sideways to a wall penetration, with one major suspect eliminated on evidence.

When testing earns its cost

Testing pays when the cost of acting on a guess is high.

A strata facing a recommendation to reclad an entire elevation on the theory that the windows leak should test before committing. If the water is actually entering at a parapet or a roof-to-wall detail, the recladding fixes nothing and the leak continues into a newly finished wall.

Testing also pays in a warranty dispute. A documented test at a stated pressure, with photographs and an observer, is evidence. An opinion from a trade with an interest in the outcome is not, and the difference matters when a claim under the 2-5-10 home warranty is being assessed.

It pays again on new work as a check before the crew leaves. Testing a sample of newly installed windows or newly sealed joints while the access is still up and the installer is still on site costs a fraction of testing them two years later, and a failure found then gets corrected under the contract instead of argued about afterwards.

Timing it

Late summer into early autumn is the practical window in Metro Vancouver. Several dry days before the test let the cladding and the cavity dry out, so any water appearing inside during the test came from the test.

Run the same test in February, after three weeks of rain, and a wet observation point is ambiguous. It may be the test. It may be Tuesday. Ambiguity is the one thing the exercise is meant to remove, so it is worth waiting for the weather rather than paying for a result nobody can rely on.

The short version

Water testing turns a leak argument into a leak fact. It works by wetting one detail at a time while everything else stays dry, which is the opposite of what rain does.

E1105 with its pressure chamber is the rigorous method and the one that reproduces wind-driven rain. AAMA 501.2 with its nozzle is faster and narrower, and it belongs on fixed joints rather than operable windows.

Preparation, particularly removing interior finishes so water in the cavity is visible, is what separates a real test from an expensive one that tells you nothing. Access is usually the largest line item, so group locations into one mobilization.

Test enough locations to support the conclusion being drawn, read the result as applying only to what was tested, and run it in a dry spell.

Done properly, it is the cheapest way to avoid paying for the wrong repair.

Sources and further reading

• ASTM International, ASTM E1105: Standard Test Method for Field Determination of Water Penetration of Installed Exterior Windows, Skylights, Doors, and Curtain Walls, by Uniform or Cyclic Static Air Pressure Difference, referenced 2026
• American Architectural Manufacturers Association, AAMA 501.2: Quality Assurance and Diagnostic Water Leakage Field Check of Installed Storefronts, Curtain Walls, and Sloped Glazing Systems, referenced 2026
• BC Housing, Building Enclosure Design Guide, water penetration testing guidance, referenced 2026
• Homeowner Protection Office, Guide to Building Enclosure Renewals, referenced 2026
• Allweathercoating field investigation records, Metro Vancouver, 2026