The Air Barrier “Passed Inspection.” The Building Still Leaked Air.

The test result said everything was fine.

Air leakage numbers came in under the required threshold. The assembly met performance criteria. The air barrier had been installed, inspected, and signed off.

And yet… occupants were complaining.

  • Drafts near windows
  • Cold spots along walls
  • HVAC running harder than expected

Nothing catastrophic. Just enough to tell you something wasn’t quite right.

So we went looking.


On Paper, This Was a Strong System

This wasn’t a cheap or outdated approach.

The project used a fluid-applied, vapor-permeable air barrier, designed to:

  • Control air movement through the wall assembly
  • Allow vapor to escape (≥10 perms)
  • Maintain continuity across transitions and penetrations
  • Meet strict air leakage requirements (≤0.04 cfm/sf at test pressure)

It’s the kind of system that, when done right, performs extremely well.

Which is why this one was so frustrating.


The Membrane Wasn’t the Problem

We opened up sections of the wall expecting to find:

  • Thin spots
  • Missed areas
  • Poor adhesion

Instead:

  • Film thickness was within range
  • Coverage was consistent
  • Adhesion looked solid

The membrane itself was doing its job.

So why was air still getting through?


The Issue Was Continuity — Not Coverage

This is the part that separates real-world performance from spec compliance.

Air barriers don’t fail because of what’s on the wall.

They fail because of what’s between everything else.


1. Transitions Were the Weak Link

The spec required:

  • Continuous connection to windows, roofing, and adjacent systems
  • Proper use of transition strips, sealants, and accessories

And those details were installed.

But not always perfectly.

We found:

  • Slight gaps where fluid membrane met transition strips
  • Inconsistent adhesion at tie-ins to window frames
  • Areas where sealant didn’t fully bond to both substrates

Nothing obvious. Nothing that would fail a visual inspection.

But air doesn’t need obvious.

It just needs continuous pathways.


2. Penetrations Were “Sealed” — But Not Integrated

Every penetration had been addressed:

  • Pipes
  • Fasteners
  • Wall ties

Sealant was used. Foam was applied.

But here’s the difference:

Sealing a penetration is not the same as integrating it into the air barrier system.

Some penetrations lacked:

  • Reinforced transition material
  • Full membrane tie-in
  • Redundant sealing

So over time:

  • Slight movement occurred
  • Materials separated microscopically
  • Air found a path

3. Substrate Prep Was Good — Not Perfect

The spec required:

  • Clean, dry, properly prepared substrates
  • Removal of contaminants
  • Smooth, continuous surfaces

All of that was done.

But in practice:

  • Some CMU joints weren’t fully flush
  • Minor surface irregularities remained
  • A few areas had borderline moisture during application

Fluid-applied systems are forgiving — more than sheet membranes.

But they still rely on:

Continuous contact + proper thickness + solid substrate

Miss one of those slightly, and performance drops — even if the install “passes.”


4. The System Passed a Test — Not Real Life

This is the uncomfortable truth.

Air barrier testing:

  • Happens under controlled conditions
  • Measures performance at a moment in time

Buildings:

  • Move
  • Expand and contract
  • Experience wind, temperature swings, and pressure differences

So a system can pass testing…

…and still develop leakage paths later.


What Changed After This Project

We didn’t change products.

We changed how we think about air barriers.


We Stopped Thinking “Coating” — Started Thinking “System”

The membrane is just one part.

The real air barrier is:

  • Membrane
  • Transitions
  • Sealants
  • Penetrations
  • Interfaces with other systems

If any one of those breaks continuity, the system fails.


We Elevated Transition Details to Priority #1

Not “important.”

Critical.

Because:

  • Walls are easy to coat
  • Details are where air leaks

We now assume:

Every transition is a potential failure point until proven otherwise.


We Treat Penetrations Like Movement Zones

Instead of just sealing:

  • We reinforce
  • We integrate
  • We allow for movement

Because static details don’t survive dynamic buildings.


We Respect Environmental Conditions More Than the Spec Does

Specs say:

  • Don’t apply to wet substrates
  • Follow temperature guidelines

Reality says:

  • “Almost dry” isn’t dry
  • Weather changes faster than schedules
  • Cure conditions matter long after application

The Takeaway

Fluid-applied air barriers are incredibly effective.

They:

  • Provide seamless coverage
  • Handle complex geometry well
  • Allow vapor to escape while stopping air

But they depend on one thing above all:

Continuity.

Not 95% continuity. Not “good enough.”

Continuous. Everywhere.


The Lesson That Sticks

You can coat every square inch of a wall perfectly…

…and still have an air leak.

Because air doesn’t go through the membrane.

It goes around it.

And if you don’t treat every transition like it matters — it will find a way.