Mastic, not tape
A thick paste applied over the joint, reinforced with mesh where the gap is wide. It stays flexible, it does not dry out and release, and it is the reason sealed joints stay sealed.
Duct Sealing That Starts With a Leakage Measurement
A calibrated fan gives a number before the work and a number after it, which is the only way anyone can tell whether sealing achieved anything.
- Leakage measured before and after
- Leaks to outside separated from total
- Combustion safety rechecked after work
- Both numbers left with the homeowner
What sealing can and cannot do
It can recover air that was being delivered into an attic or crawl space, reduce the load on the equipment, and often resolve a room that was running short because its branch was leaking.
It cannot make an undersized branch bigger, uncrush a compressed run, or add return capacity a system never had. Those are distribution problems and they get identified separately.
It can also change what the blower is working against, which is why static pressure is measured on both sides of the job.
Duct leakage is unusual among building problems in that it is completely invisible, entirely measurable, and concentrated in a handful of predictable locations. A household never sees it. A technician with a duct testing fan quantifies it in under an hour. And the bulk of it is almost always at the plenum seams, the branch takeoffs and the boots, rather than distributed evenly along the runs.
That combination is what makes sealing worth doing properly rather than approximately. Work that concentrates on the few locations responsible for most of the loss produces a measurable change. Work that involves walking around a basement with a roll of tape does not, and there is no way to tell the difference afterwards unless somebody measured.
The consequence of leakage depends heavily on where the duct runs. A supply joint leaking into a conditioned basement is wasteful but modest. The same joint leaking into a vented attic is delivering heated or cooled air straight outside, and on a branch that has fully separated from its takeoff, that can be a substantial fraction of what the equipment produces.
Sealing work in Melissa, TX has been measured, targeted and measured again since 2006. Where a system also has a restriction that the leakage was masking, that gets reported rather than discovered later.
Measuring before proposing
A duct testing fan is a calibrated blower that connects to the system, usually at the return or through a removed register, with every other opening temporarily sealed. It pressurizes the duct to a reference pressure and reports the airflow required to maintain it. That airflow is the leakage rate, expressed as a volume per unit time.
Running the same test with the building itself pressurized to the same reference removes the pressure difference across any duct inside conditioned space, so those leaks stop registering. What remains is leakage to outside, which is the figure that matters most for both running cost and comfort.
Having both numbers changes the conversation. A system with substantial total leakage but very little of it to outside is a lower priority than one where nearly all of it escapes the building envelope, even if the headline figures look similar. Sealing gets recommended or not on that basis rather than on the assumption that all leakage is equally worth chasing.
Finding the individual leaks
The fan gives a total. Locating the contributors is done by hand with the system pressurized: feeling along seams, using smoke at suspected joints, and inspecting the accessible runs directly. Thermal imaging helps in some situations, particularly where duct runs behind a surface and the temperature difference is enough to show through.
Most of what gets found is unspectacular. A plenum seam that was screwed and never sealed. A takeoff with a visible gap around it. A boot that was fitted to the opening with a nominal wrap of tape a long time ago. Occasionally something dramatic turns up, such as a branch lying completely detached in an attic, and that single finding can account for more than everything else combined.
| Location | Why it leaks | How it gets sealed |
|---|---|---|
| Supply and return plenums | Large seams, high pressure, frequently assembled without sealant | Mastic with mesh reinforcement over every seam and corner |
| Branch takeoffs | Cut opening larger than the fitting, minimal original sealing | Mechanical fastening, then mastic around the full collar |
| Boots at registers | Gap between the boot and the floor or ceiling opening | Mastic or foil tape between boot and building material |
| Flexible duct connections | Inner liner not banded properly, outer jacket left open | Reband the liner, mastic the collar, close the vapor jacket |
| Disconnected branches | Never banded, or pulled apart by weight over time | Reconnect, support properly, then seal |
| Equipment cabinet and filter rack | Access panels and rack slots open directly to the return | Gasket the panels and close the rack slot |
The safety reason some leaks come first
Return side leakage inside a mechanical room, a garage or a basement containing an atmospherically vented combustion appliance is treated as a priority regardless of how much it contributes to the total.
The mechanism is straightforward. A leaking return draws air out of the space it sits in. That space becomes slightly negative relative to outdoors. An atmospherically vented water heater or furnace in the same space relies on a gentle natural draft up its flue, and a negative pressure in the room can be enough to overcome it, pulling combustion products back into the building instead.
This is not theoretical and it is exactly why any sealing work in such a space is followed by a combustion safety check. The appliance is tested under worst case conditions, meaning with exhaust fans running and the building configured to maximize depressurization, and draft is confirmed at the flue.
Where the check shows a problem, it gets addressed before the job is signed off. Improving duct tightness is not worth creating a venting hazard.
Mechanical first
Screws or bands hold the joint together; sealant closes it. A joint relying on sealant for structural connection will open again as the duct moves with temperature.
Before and after numbers
The leakage test is repeated once the work is complete. Without the second number, nobody including the contractor knows what was achieved.
Static pressure both sides
Sealing changes system resistance. Measuring before and after catches the case where leakage had been masking an underlying restriction.
When sealing is the wrong recommendation
There are systems where it is not worth doing, and saying so is part of an honest survey. A system that tests tight already has nothing to gain. A system where nearly all the duct runs inside conditioned space has far less to gain than the headline leakage figure suggests.
There are also systems where sealing would be premature. If the distribution is about to be substantially reworked because a branch is undersized or a trunk was reduced during a renovation, sealing first means paying to seal duct that is coming out. Where an equipment replacement is imminent, the plenums will be opened anyway.
In each of those cases the sensible sequence is stated plainly, even where it defers the work or removes it entirely.
A sealing job with no measurement at either end is unverifiable by definition. Ask for the before and after numbers, and treat their absence as the answer to how the work was done.
What the improvement actually buys
The clearest gain is air arriving where it was intended. A branch that was losing a significant share of its flow into a crawl space starts delivering that air to the room at its end, and a bedroom that never kept up can become ordinary. That is the change households actually notice.
The second gain is load. Air escaping a supply duct into an attic has to be replaced, and it is replaced by unconditioned air drawn in through the building envelope. The equipment then conditions that too. Closing the leak removes both halves of that cycle, which is why duct leakage costs more than the volume of the leak alone suggests.
The third is dust and air quality, which applies mainly to return side leakage. A return drawing air from an attic, a crawl space or a garage is pulling whatever is in those spaces into the distribution system and then into the living space. Sealing the return stops that, and it is often the reason a household with a persistent dust complaint sees a change.
What nobody should expect is a specific percentage saving, because that depends on the leakage rate, where the duct runs, the climate and how the building is used. The honest deliverable is a measured before figure and a measured after figure.
How the visit runs
Baseline the system
Static pressure supply and return, then a duct leakage test giving total leakage and, where the equipment allows, leakage to outside.
Locate the contributors
Accessible runs inspected with the system pressurized, using touch, smoke and thermal imaging as appropriate to find the individual sites.
Report and quote
Findings shown, the leak sites listed, and the work quoted as one flat figure with anything inaccessible identified rather than glossed over.
Reconnect and support
Anything disconnected is reconnected and properly supported first, since sealing a joint that is about to pull apart again achieves nothing.
Seal the concentrations
Plenums, takeoffs and boots first, because those account for most of the measurable loss, then remaining accessible joints.
Close the equipment cabinet
Access panel gaskets and the filter rack slot, which sit directly on the return and are frequently the easiest gain on the whole job.
Retest
Leakage measured again and static pressure retaken, so the change is quantified rather than asserted.
Combustion safety check
Where any atmospherically vented appliance shares a space with the ductwork, draft is confirmed under worst case depressurization before the job closes.
- Leakage measured before any sealing is proposed
- Leaks to outside separated from leaks inside where possible
- Mechanical fastening before sealant, mastic rather than tape
- Leakage and static pressure both retested at the end
- Combustion safety verified where a vented appliance shares the space
Leakage questions
What sealing achieves and how it gets verified
Sealing is easy to claim and hard to see, which is exactly why the measurements matter more here than on most work.
With a calibrated fan connected to the duct system while every register and return grille is temporarily sealed. The fan pressurizes the ductwork to a reference pressure, and the airflow it has to deliver to hold that pressure is the leakage rate. It is a number, not an estimate.
The same equipment can be used to separate total leakage from leakage to outside, which matters because a leak into a conditioned basement wastes far less than an identical leak into a vented attic. Knowing which you have determines whether sealing is worth doing and where to concentrate it.
Concentrated in a small number of places rather than spread evenly. The plenum seams where the duct meets the equipment, the takeoffs where branches leave the trunk, the boots where duct meets a floor or ceiling opening, and any joint that was fastened but never sealed.
Branches that have come completely adrift in an attic or crawl space are the extreme case and they are not rare. A supply branch delivering conditioned air directly into an attic is invisible from inside the house and can account for a large share of the total on its own.
No, and the name is one of the trade's long standing jokes. The adhesive on conventional duct tape dries out under temperature cycling and the tape releases, typically within a few years, leaving the joint open again.
Sealing is done with mastic, a thick paste applied over the joint and reinforced with mesh where the gap is wide, or with foil backed tape rated for the purpose where mastic is impractical. Mechanical fastening comes first in either case: sealant closes a joint, it does not hold one together.
Not entirely, though they are much less costly than leaks to outside. Air escaping a supply joint in a conditioned basement is still heated or cooled air that is not arriving where it was meant to, so the room at the end of that branch still runs short.
There is also a pressure effect. Leakage on the return side inside a mechanical room or a garage pulls air from that space into the system, which is worth attention both for comfort and, where a combustion appliance shares the space, for safety.
Sometimes it resolves it completely, particularly where the branch serving that room was leaking badly or had partially disconnected. Air that was escaping into a crawl space starts arriving at the register instead.
Where the branch is undersized or crushed rather than leaking, sealing improves the system generally but will not fix that room. The distinction is established during the survey, so nobody pays for sealing expecting it to solve a sizing problem.
Yes, and it is the reason return side leakage in certain locations gets priority. A return leaking inside a space that contains an atmospherically vented combustion appliance can depressurize that space and pull flue gases back down the vent instead of letting them rise.
That is a carbon monoxide risk rather than an efficiency one. Any sealing work in a mechanical room with a combustion appliance includes a combustion safety check afterwards, confirming the appliance still drafts correctly under worst case conditions.
The duct itself, no. But sealing a leaky system changes its total resistance, which changes what the blower is working against. A system that had been getting away with an undersized return partly because it was drawing air through leaks may show its true static pressure once those leaks close.
This is why static pressure is measured before and after. If sealing reveals an underlying restriction, that gets reported rather than left to surface later as a noise complaint or a frozen coil.
Most residential systems are a day, sometimes less. The time goes into access rather than into the sealing itself, since a good deal of the work happens in crawl spaces, attics and behind equipment.
Where the system is largely concealed in finished ceilings, the accessible portion is sealed and the remainder is reported honestly rather than claimed. Sealing what can be reached in a typical home addresses the large majority of measurable leakage because the worst offenders are concentrated at the plenums and takeoffs.
It exists and it has a place, particularly where duct is largely inaccessible behind finished surfaces. A sealant is atomized inside the pressurized duct and deposits at the leak edges.
It is not a substitute for hand sealing where access exists, because mastic applied directly to a plenum seam or a disconnected takeoff is more durable and addresses the large openings that aerosol methods handle least well. Where it is appropriate it gets proposed on that basis rather than as a default.
Work that sits alongside sealing
Leakage, sizing and air quality are three views of the same distribution system.
If the system tests tight, the test says so plainly and there is nothing here worth selling you. If nearly all of the duct runs inside conditioned space, that gets said too, because the case for sealing is much weaker there than the raw leakage figure suggests.
Testing and sealing across Melissa, TX, booked through the same dispatch desk as repair and installation work. If you are already planning an equipment replacement, mention it, because the plenums come apart during that job anyway and sealing them then costs considerably less than sealing them twice.