Why $300 of Hand Caulking Still Fails the Blower Door

On a 50-unit project in Springfield, Massachusetts, the crew caulked every framing gap by hand. The cost ran about $300 a home. The blower door test failed anyway. That outcome surprises people, because the work was done and it was done conscientiously. Every gap the crew could see got a bead of caulk in it. The problem is that caulk shrinks as it dries, and almost half of new homes fail their first test for the same reason. The $300 was never the expensive part of that job.

The Real Test Result

The number that gets quoted for hand air sealing is the material plus the labor to apply it. On this project that was about $300 a home, which sounds reasonable against everything else in the budget. It is reasonable, right up until the test does not pass.

What the $300 does not carry is the second half of the cycle. When a blower door fails, someone has to go find the leaks, seal them again, and pay for a retest. If the failure is caught late, that means opening up work that was already closed. Meanwhile the trades stacked behind the test are waiting, and on a 50-unit building the waiting is not happening once.

That is the part worth sitting with. The crew did not do a bad job. They sealed what was in front of them with the material they were handed. The method is what failed, and a method that fails almost half the time is not a line item, it is a schedule risk that happens to have a small invoice attached to it.

The Shrinkage Problem

Caulk goes in wet and cures by giving up what makes it wet. Losing that volume is what curing is. A bead that completely filled a gap on the day it was tooled does not completely fill it a week later, and the gap it was bridging is not holding still either, because framing lumber keeps drying and moving after the crew leaves. Shrinking material across a widening joint separates from at least one face. Air only needs one.

The second half of the problem is reach. Hand caulking seals the gaps a person can see and get a gun into. The leakage that actually fails a test is spread across places nobody is looking, which means top plates, rim and band joists, electrical boxes, plumbing penetrations, sheathing seams, the framing behind a tub, and the ceiling plane above a soffit. Those add up quietly, and none of them announce themselves during the walk.

The statistic is the tell. Almost half of new homes fail their first test, and they are not failing because half the framing crews in the country are careless. They are failing because visually directed sealing with a shrinking material is being asked to hit a measured number, and it is a poor tool for that job.

Aerosol sealing inverts the search. The unit gets pressurized and a sealant fog is released inside it, so the air leaving the enclosure carries the particles to every leak at once and deposits them at the edges of the gap until it closes. The leaks find themselves. Nobody has to know where they are. Just as important, the operator is watching the leakage number fall in real time and works to the target, so the result is a measured number rather than a hope that the walk was thorough.

In Massachusetts that distinction gets sharper, because multifamily and townhouse projects are tested for compartmentalization unit by unit, in CFM50 per square foot of enclosure rather than as a single whole building number. On a 50-unit building that is 50 separate chances to fail rather than one, and a method that passes most of the time is going to find the units where it does not.

This is also where the downstream money sits. Air leakage is an input to the energy model like any other. When the sealing number is reliable and tight, the model has margin to give back, and that margin usually gets spent on insulation thickness and glazing you no longer have to buy. When the sealing number is a coin flip, nobody can spend it in advance, so the envelope and window packages carry the risk instead.

When Hand Caulking Still Works

When it is the backup rather than the strategy. Sealing large visible gaps by hand before an aerosol pass is normal and useful. Very large openings should be closed with foam or blocking regardless, because no method should be asked to bridge a hole it was not designed for. The failure mode above is hand caulking carrying the whole target by itself.

When the unit count is too low to justify mobilization. Aerosol equipment has a setup cost that gets divided across the units it treats. On a single home or a two unit building, that division works against you, and a careful hand pass plus a generous margin in the model can be the lower cost route.

When the target is base code rather than a stretch number. The further the target sits from code minimum, the less tolerance there is for a method that lands in a range. Where the required number is loose, a thorough hand pass has room to be imperfect and still pass.

When the assembly is genuinely simple. A compact rectangle with few penetrations, a simple roof plane, and no complicated intersections gives hand sealing a real chance, because the list of places to look is short enough to actually finish.

When the schedule cannot accommodate the sequencing. Aerosol sealing has to happen at a specific point, with the enclosure at the right stage and openings masked. If the project cannot hold that window, forcing it in creates coordination problems that cost more than the retest risk it was meant to remove.

The Decision Framework

Work backward from the number you have to hit. Find out what your target actually is and how it gets measured, which in Massachusetts multifamily means per unit rather than per building. Then ask an honest question about the method in the specification, which is whether it hits that number repeatably or merely usually. Those are different products, and only one of them can be relied on before the fact.

Then price the full cycle instead of the bead. The comparison is not $300 a home against an aerosol number. It is $300 a home plus the odds of a retest, the rework behind it, and the days the schedule loses, against a number that arrives with the target already met. Once the second column is on the page, the arithmetic on a 50-unit building usually stops being close.

Where it gets genuinely interesting is what you do with the margin afterward. A tight, predictable enclosure is the input that lets the rest of the envelope get less expensive, and how much it gives back depends on your climate zone, your window package, your mechanical system, and how much headroom the rest of the model already has. That is why the answer moves from building to building. Springfield learned what its method was worth by failing a test, which is the expensive way to find out. Running the number on your own project first is the other way.