How the two differ in kind
Noise from mechanical plant reaches a room by two fundamentally different routes, and treating the wrong one is expensive and ineffective.
AIRBORNE noise travels in the airstream, down the duct, and out of a grille — or radiates from the plant into the plant room and through its walls. That is what attenuators, lagging and acoustic construction address, and it is the route people think of first because there is a visible path.
STRUCTURE-BORNE noise does not use the air at all. A fan, pump, compressor or chiller vibrates; if it is hard-mounted to a slab or a frame, that vibration goes straight into the structure, travels through it with very little attenuation, and radiates from surfaces elsewhere in the building. The symptom is distinctive: a hum or rumble that seems to come from the walls and floor rather than from any opening, that is audible in rooms nowhere near the plant, and that is completely unaffected by anything done inside the ductwork.
The tell is location. If the noise is loudest at a grille, it is airborne and duct treatment applies. If it is present in a room with no ventilation opening at all, or in a room on another floor, it is arriving through the structure and an attenuator will do nothing.
The remedy is ISOLATION — springs, rubber mounts, an inertia base, isolated hangers, flexible connections at the duct and pipe — and it has a property worth understanding before selecting anything. An isolator works by having a natural frequency well below the frequency of the disturbance, and how far below is governed by its STATIC DEFLECTION: how much it compresses under the load it carries. An isolator that barely compresses is barely isolating. Worse, an isolator whose natural frequency lands near the equipment's running speed AMPLIFIES the vibration instead of reducing it, which is why the selection depends on knowing the operating speed and the supported weight rather than on fitting something soft.
The factors that actually differ
Show
| Isolating the plant | Silencing the duct | |
|---|---|---|
| Path it addresses | Vibration entering the building structure from the equipment. | Sound travelling inside the duct toward a grille. |
| Symptom it fixes | A hum or rumble from the walls and floor, audible in rooms with no opening and on other floors. | Noise at the grille, loudest in the room being served. |
| The governing quantity | Static deflection under load, which sets the isolator's natural frequency relative to the running speed. | Insertion loss per frequency band, and the pressure drop it costs. |
| How it can backfire | An isolator near resonance with the running speed AMPLIFIES rather than isolates. | An oversized attenuator adds pressure drop, so the fan works harder and becomes a noise source. |
| What defeats it | A single rigid bridge — a pipe, a conduit, a drain line or a duct connection hard-fixed across the isolation. | Break-out upstream of the attenuator, and any structure-borne path. |
| Where it goes | Under the equipment, in the hangers, and at every service connection crossing to the structure. | In the duct run, near the source, with straight duct either side. |
| Effect on the other problem | None on airborne noise in the duct. | None on structure-borne noise at all. |
| Flexible connections | Essential at duct and pipe connections, or the service becomes the bridge that defeats the mounts. | Not applicable. |
| Low frequency | This is where it earns its place — structure-borne rumble is characteristically low-frequency. | Weak at low frequencies, which is why fan rumble resists attenuation. |
| When to suspect it | Noise in rooms with no duct, on other floors, or that continues when grilles are blocked. | Noise that clearly comes out of the grille and falls away across the room. |
Which one, and when
Choose isolating the plant when…
- The noise is audible in rooms with no ventilation opening, or on a different floor from the plant.
- It is a low-frequency hum or rumble that seems to come from surfaces rather than from an opening.
- Plant is being installed on a structural floor above or adjacent to occupied space.
- Duct treatment has already been fitted and the complaint has not changed.
Choose silencing the duct when…
- The noise is loudest at the grille and falls away with distance from it.
- It is mid- or high-frequency hiss or broadband noise rather than a low rumble.
- A fan serves a sensitive room directly and the duct-borne path is the obvious route.
- There is space in the run and static pressure available for the attenuator's drop.
Now run your own numbers
This page holds no prices on purpose — a national average is wrong for almost every real project. Quantify both options with your dimensions and your local quotes.
Frequently asked questions
- How do I tell structure-borne from airborne noise?
- By where it can be heard. Airborne noise from a ventilation system needs a path through the air, so it is loudest at the grille and diminishes across the room; block or close the grille and it changes. Structure-borne noise needs no opening at all, so the tell is hearing it in a room that has no ventilation terminal, in a room on another floor, or in a space diagonally adjacent to the plant with no duct between them — and it does not change when grilles are covered. Its character is usually distinctive as well: a low hum or rumble that seems to emanate from the walls and floor rather than from a point, sometimes with a tonal quality matching the equipment's running speed. If the complaint fits that description, nothing done inside the ductwork will resolve it.
- Why does static deflection decide isolator performance?
- Because it sets the isolator's natural frequency, and isolation only happens when that frequency is well below the frequency of the disturbance. A softer isolator compresses more under the same load, which lowers its natural frequency and increases the separation from the running speed — so static deflection is a direct proxy for how much isolation you get. An isolator that barely compresses under its load has a high natural frequency and is doing very little. The selection therefore depends on two numbers: the weight each mount carries, which determines the deflection a given spring will give, and the equipment's lowest disturbing frequency, usually its running speed. Manufacturers publish the deflection required for a given isolation efficiency at a given frequency, and the mount is chosen from that rather than by feel.
- Can an isolator make vibration worse?
- Yes, and it is a real failure rather than a theoretical one. Every spring-mass system has a resonant frequency, and near it the response is AMPLIFIED rather than reduced — so an isolator whose natural frequency happens to sit close to the equipment's running speed transmits more vibration into the structure than a rigid mount would. That is why the selection depends on knowing the operating speed, and why variable-speed equipment needs care: a unit that runs across a range may pass through the resonance at some speeds. It is also why every machine passes through its resonance during start-up and run-down, which is normal and brief, and why snubbers are fitted on some installations to limit movement during that transition. Fitting something soft because soft sounds better is exactly how this goes wrong.
- What defeats a correctly chosen isolator?
- A single rigid connection across it, and this is the most common reason an isolated installation still transmits. Equipment sitting perfectly on springs but connected to rigid duct, rigid pipework, rigid conduit or a hard-piped condensate drain has a direct path into the structure through every one of those — and the springs are irrelevant, because vibration takes the stiff route. The fix is flexible connections on every service crossing the isolation boundary: flexible duct connectors, flexible pipe connectors or loops, flexible conduit, and a drain arrangement that does not bridge. The same applies to anything touching the equipment incidentally — a wall it leans against, a stiffener that was added, a pipe support fixed to both the frame and the slab. Commissioning should include walking round and checking that nothing rigid crosses.
- What is an inertia base for?
- Adding mass beneath the equipment so the assembly is heavier and more stable, which does several things at once. It lowers the centre of gravity and resists the rocking that a lightweight machine on soft springs is prone to, particularly at start-up and where the equipment produces a thrust — a fan discharging into a duct pushes back. It increases the mass on the isolators, which allows a given spring to deflect more and therefore isolate better. And it evens out the load distribution across the mounts. Typically a concrete-filled steel frame, it is used where the equipment is light relative to its disturbing force, where the required deflection is large, or where the machine's own frame is not stiff enough to sit on point mounts without flexing.
- Do hangers need isolating too?
- Yes, and suspended equipment and services are a frequently missed path. A fan, an air handling unit or a length of duct hung on plain threaded rod transmits straight into the slab above, which then radiates into the room below — and the room below is often the one complaining. Isolation hangers, containing a spring, a rubber element, or both, break that path, and they are selected on the same basis as floor mounts: the load carried and the deflection required. Two details matter. The rod must not touch the hanger housing anywhere, since a rod resting against the edge of its clearance hole is a rigid bridge. And the isolation has to continue along the run: an isolated unit feeding duct that is hung rigidly a few metres away is still connected to the structure.
- Which should be done first?
- Neither, until the path has been identified — which is the whole argument of this page. Both treatments are expensive, both are hard to retrofit, and each does nothing for the other's problem, so the sequence starts with diagnosis: where is the noise audible, does it change when grilles are closed, is it present in rooms with no duct, and does its character suggest low-frequency structure-borne rumble or mid-frequency airborne hiss. In design, the answer is usually both, because plant produces both kinds of noise — isolation under the equipment, flexible connections at every service, and attenuation in the duct near the source. In a complaint on an existing installation, doing the diagnosis first is what avoids fitting an attenuator to a problem that was never in the duct.
- Does the plant room construction matter?
- A great deal, and it is the third path alongside the two this page compares. Noise radiating from the equipment into the plant room has to be contained by the room itself, which means walls and a floor-ceiling construction with adequate mass and, critically, sealing — because a plant room with a well-built wall and an unsealed door, an open duct penetration or a gap around a pipe leaks sound through the gap regardless of the wall. Doors are usually the weakest element and acoustic doors with proper seals are the standard answer. The floor matters where there is occupied space below: mass in the slab helps, but a heavy slab still transmits whatever is fed into it rigidly, which is why isolation under the equipment remains the primary control rather than the room's construction.