Why a Door or a Duct Is Usually the Weakest Element

Why this matters

A partition is rarely one thing. It is a wall, plus whatever had to open, plus whatever had to pass through. The wall is the part that gets specified, argued about and upgraded, and it is usually not the part that decides the outcome, because a wall's performance comes from mass and construction while a door's comes from a perimeter seal and a duct's comes from an air path that does not meaningfully have a wall in it at all. Rank the elements correctly and the fix is obvious and usually cheap in labor hours. Rank them by which one is easiest to change and you can spend a whole scope raising a number that was already irrelevant.

The metric that ranks elements honestly

Do not rank by rating. A 35 dB door and a 50 dB wall do not tell you which one to fix, because they act over different areas. Rank by equivalent open area, which is area multiplied by transmission coefficient. That product is what sums across elements, so it is the only ranking that predicts what happens when you change one.

equivalent open area = element area x 10^(-TL/10)

Add those products, divide by total area, convert back with 10 log10 of the reciprocal, and you have the composite. The small-hole card owns that arithmetic; what this card owns is which elements land where in the ranking and why.

What each element's number is actually a property of

That table is the category error this group is about. A wall's number is a property of construction. A door's number is a property of a seal, and a seal is a consumable that ages, gets painted over, gets adjusted for a sticking latch, and gets removed by a carpenter fixing a complaint about drag. A duct's number is a property of a path, and a path can be added to by anyone who cuts in a takeoff two years later.

Why the door is usually second and the duct usually first

A door assembly is tested to ASTM E90 as an operable partition and classified per ASTM E413, in the editions the specification cites, with every seal in its tested condition. Two things follow. Removing the automatic drop seal does not reduce the assembly by a few points; it converts the bottom of the door into an open slot, and an open slot is governed by area fraction rather than by mass. And a door that does not latch tight leaves the compression seals uncompressed all the way around, which is the same failure repeated on four edges.

A duct is worse because it is a two-way waveguide with no barrier in the path. Sound entering the duct in one room travels the duct and exits in the other, essentially unimpeded except by whatever lining, bends and end reflections are in the way. It has a second route as well: energy inside the duct radiates out through the casing on the receiving side, which is breakout, and breakout has its own card. If the casing radiates more than the in-duct path delivers to the outlet, the silencer you added is no longer the governing number.

Which puts the elements in a usual order, and usual is doing work in that sentence: the transfer opening or duct crossing first, the door second, the wall a distant third. It is usual rather than universal, and the way to know is to compute the equivalent open areas rather than to assume the order.

Worked example: one wall, three elements

An office wall between a conference room and a corridor. Gross area 12 ft by 10 ft = 120 ft2. All values are field figures, unweighted, per octave band basis where stated, and single-number where the element is rated that way.

• Door, 3 ft by 7 ft = 21 ft2, tested door assembly rated 35 dB with seals in their tested condition.
• Transfer duct through the wall, 12 in by 8 in = 0.667 ft2, with a lined transfer element.
• Wall field, 120 - 21 - 0.667 = 98.333 ft2, performing at 50 dB.

Correction, printed: the duct's number is a re-basing, not a rating. The lined transfer element is published with an insertion loss, which is measured against a reference condition, the same duct run without it. Insertion loss is not the transmission loss of the opening in the wall. Using the 15 dB insertion loss as the path's transmission loss in a composite is an approximation, and it holds only while the duct casing's breakout on the receiving side is higher than what the in-duct path delivers. If the casing radiates into the receiving room, the casing sets the floor and this number stops governing. Stated here as an approximation, and it is the one figure in this example that would need a real duct-borne and breakout analysis on a live job.

Correction, printed: areas are net. The wall term uses 98.333 ft2, not 120. Charging the gross area to the wall counts the door and the duct twice.

Composite: 10 log10 (1 / 2.39e-4) = 36.2 dB. A wall specified at 50 is performing at 36.

Fix the wall. Take it from 50 to 60, which is a real assembly change. New total equivalent open area 0.0278 ft2, composite 36.4 dB. Gained 0.2 dB. Ten decibels of wall bought two tenths.

Fix the duct instead. Take the transfer element from 15 to 30 dB. New total 0.00829 ft2, composite 41.6 dB. Gained 5.4 dB, which is past the roughly 3 dB most people report as a change in a steady broadband sound.

Ceiling check, printed. Before the duct fix, everything other than the duct is 0.00762 ft2. The ceiling on perfecting the duct is 10 log10 (0.0287 / 0.00762) = 5.8 dB. Achieved 5.4 dB, under its own ceiling as it must be.

And where it goes next. After the duct fix the door holds 80% of the remaining equivalent open area, and its own ceiling is 10 log10 (0.00829 / 0.00165) = 7.0 dB. The wall is still third. It was third before the job started and it is third after.

Sibling-rule check, printed. Elements summed on area-weighted transmission coefficient rather than by subtracting decibels, per the small-hole card: yes. Areas taken net so no element is counted twice: yes. Every gain quoted against the ceiling that caps it, per the flanking card: yes. The duct figure labelled as a re-based approximation with the condition it holds under, rather than used as if it were a measured transmission loss: yes. No rounding taken in the flattering direction: the gains are quoted below their ceilings and the duct approximation is flagged as the optimistic one.

What would change the ranking

A door with no undercut and a functioning drop seal, in a wall that has no services crossing it, can drop below the wall in the ranking. That is the whole argument for specifying the seal set as part of the door assembly rather than as hardware.

A duct crossing that is long, lined and offset rather than a straight transfer opening changes the in-duct path enough to move it out of first place, at a cost in static pressure that a sibling card owns. A straight sleeve is the worst case and is common because it is what fits.

A masonry or concrete wall raises the wall's own performance and therefore lowers its share further, which makes the openings dominate even harder. Heavier walls do not reduce the importance of the openings; they increase it.

Low frequency inverts the door case. A door's advantage over a hole comes from its mass, and mass buys least at low frequency, so a bass complaint through a door is a different problem from a speech complaint through the same door.

Working on these elements without creating a second problem

• Rated doors. Do not defeat a self-closing or self-latching device, do not add a gasket, sweep or seal that is not part of the door assembly's listing, and do not increase the undercut on a rated door. NFPA 80 applies in the edition your authority having jurisdiction adopted, binds the building owner, and reaches you through the permit; 29 CFR 1910.36 and 1910.37 cover exit routes as a federal duty owed to your employees.
• Reaching into a duct. Shut down the fan at its disconnect and lock and tag it before any hand or tool enters the airstream, per 29 CFR 1910.147; equipment on building-automation control starts on a schedule with nobody at the panel.
• Airflow you just restricted. Adding a silencer, lining or offset raises static pressure. Verify airflow after the change, and never restrict a duct serving a fuel-fired appliance or a required exhaust while it can operate, because losing that airflow can spill combustion products into the space or lose capture at a hood. Where a duct penetrates a fire-resistance-rated assembly, a fire damper may be required and it may not be omitted, obstructed or held open; the mechanical code as adopted governs, and a sibling card covers the difference between fire and smoke dampers.
• Removing existing liner. Cutting or pulling fibrous duct liner releases glass fibre and whatever has settled in the system. Use dust containment and a HEPA vacuum, and use respiratory protection only under a written program meeting 29 CFR 1910.134. A comfort mask is not protection for this route.

References

• ASTM E90 and ASTM E413 for laboratory rating of walls and operable assemblies, in the editions the specification cites; consensus standards bind through that reference rather than on their own
• NFPA 80, in the edition adopted by the authority having jurisdiction; 29 CFR 1910.36 and 1910.37 for exit routes
• 29 CFR 1910.147 for control of hazardous energy; 29 CFR 1910.134 for respiratory protection programs
• See related: What a Small Hole Does to a Large Wall; What Flanking Is and Why It Decides the Outcome; Why Duct-Borne Noise and Duct Breakout Are Different Faults; What a Duct Silencer Costs You in Pressure