Insertion loss and how to specify it

Insertion loss in acoustics is the reduction in sound level at a receiver achieved by inserting something into the transmission path, measured as the difference between the level before and the level after the element is fitted. An attenuator, an acoustic louvre, an enclosure or a barrier each has an insertion loss. It is a property of the whole arrangement rather than of the component alone, which is why the same silencer can deliver 18 dB in one duct and 11 dB in another.

The number on the data sheet is a starting point. The number that matters is the one in your spectrum, at your airflow, in your duct.

What insertion loss measures

Three terms get used interchangeably and should not be.

Insertion loss is the difference in level at a fixed receiver position caused by inserting the element, with everything else unchanged. Sound reduction index, quoted as R or Rw, is a laboratory property of a partition describing how much airborne sound it stops. Attenuation is a loose word covering any reduction, including the losses from distance and duct end reflection that nobody installed.

For ducted plant there is a further distinction. Static insertion loss is measured without airflow. Dynamic insertion loss accounts for air moving through the unit, which changes the result and introduces regenerated noise from the airflow across the splitters. Specify the dynamic figure for anything carrying air.

Like the vertical sky component in daylight work, insertion loss only means something once the conditions of measurement are stated. Quote it without them and you have quoted a marketing figure.

Why a single figure is not enough

Plant noise is not flat. An axial fan typically pushes energy into the low and mid frequencies with a tonal component at blade pass frequency. A chiller adds compressor tones. An extract system serving a commercial kitchen carries broadband flow noise with a low-frequency rumble from the fan.

Attenuators are poor performers at low frequency and good at high frequency. A splitter silencer that delivers 30 dB at 2 kHz might deliver 6 dB at 63 Hz. If the problem is a 63 Hz tone through a bedroom wall at night, a headline dB(A) rating tells you nothing useful, and the installation will fail its commissioning test while nominally meeting the specification.

The fix is arithmetic rather than judgement. Take the source sound power level by octave band, subtract the insertion loss by octave band, add the regenerated noise, apply the path losses, and recombine. Then apply any character correction for tonality under BS 4142:2014+A1:2019, because a residual tone attracts a penalty that wipes out several decibels of hard-won attenuation.

Where octave band data comes from

Manufacturers publish insertion loss by octave band from 63 Hz to 8 kHz, measured to the standard test methods for ducted silencers. Ask which standard, and ask whether the figures are static or dynamic.

Two checks are worth making on any data sheet. First, whether the test duct dimensions resemble your installation, because insertion loss scales with the ratio of open area to absorptive surface and a silencer tested in a small duct will underperform in a larger one. Second, whether the regenerated sound power level is published at your design face velocity, because it is often quoted at a velocity well below the one you will run.

Where no reliable data exists, and on bespoke enclosures this is common, the honest position is to specify a performance requirement and require the supplier to demonstrate compliance by test rather than to assume a figure.

Pressure drop as the real design constraint

Every attenuator resists airflow, and that resistance is the constraint that decides how much attenuation a system can carry.

Adding a splitter silencer raises the static pressure the fan has to overcome. The fan is then selected at a higher duty, runs faster, and produces more sound power at source. Push far enough and the extra noise from the fan offsets a meaningful part of the attenuation gained, while the energy consumption rises for the life of the building.

Regenerated noise sets a floor beneath the whole exercise. Air moving across splitters generates its own sound, and once that regenerated level approaches the attenuated source level there is no benefit in a longer unit. A silencer cannot reduce noise below the noise it makes.

The design moves that avoid the trap are mostly geometric: lower face velocity through a larger cross-section, bends and lined plenums doing part of the work, and choosing quieter plant at the outset. Selecting a fan one size larger running slower is usually the single most effective acoustic decision on a mechanical services package, and it is made by the mechanical engineer months before the acoustician is asked to fix anything.

Enclosures, louvres and lined plenums

Each element behaves differently and is misapplied in a predictable way.

• Splitter attenuators give the highest insertion loss per metre in a duct, at the cost of pressure drop and length, and need roughly one to three metres of straight run
• Acoustic louvres suit an opening in a facade or screen where air must pass freely, and deliver substantially less insertion loss than a splitter unit of the same depth
• Lined plenums give broadband performance including useful low-frequency attenuation, and they need volume rather than length
• Enclosures around a whole item of plant can perform well until the ventilation openings are added, at which point those openings govern the result
• Barriers and screens work by breaking line of sight and rarely deliver more than modest reductions, with diffraction over the top setting the limit

The recurring failure is flanking. An enclosure with a high panel specification and an unsealed cable penetration performs at the level of the penetration. Detail the openings, the door seals and the service entries, or the panels are decoration.

Verifying performance after installation

Insertion loss is almost never measured directly on site, because you cannot remove the silencer and remeasure. Verification is done instead at the assessment location with the plant switched on and off, which gives the specific sound level and the residual, and the comparison is then made against the background sound level under BS 4142.

Build the verification into the programme. A commissioning test at night, with the plant at full duty and with a sound level meter at the position named in the planning condition, is what discharges the condition. Testing during the day against a high daytime background proves nothing and will be rejected.

If the result fails, the diagnostic question is which octave band is responsible. A failure at 63 Hz points at the fan and the structure-borne path. A failure across the mid frequencies usually points at flanking or a duct break-out rather than at the silencer. A noise impact assessment that recorded the predicted band-by-band levels gives you that comparison. One that recorded a single overall figure does not.

Writing it into a specification

Specify performance and the conditions of measurement, never a product.

1. State the required dynamic insertion loss by octave band from 63 Hz to 8 kHz at the design airflow.
2. State the maximum regenerated sound power level, by octave band, at the design face velocity.
3. State the maximum permitted pressure drop across the unit at design airflow.
4. Require data measured to the recognised test method for ducted silencers, with the test report supplied.
5. State the overall performance requirement at the assessment position, expressed as a rating level relative to the measured background sound level.
6. Require post-installation verification by a competent acoustician, with the method and timing agreed in advance.
7. Make any acoustic screening, plenum or enclosure a separate specified item rather than leaving it to the installer.

Point five is what the planning condition actually cares about. Points one to four are what lets the supplier price it, and what gives you somewhere to stand if the commissioning test fails. A ventilation and extraction statement drawing those requirements together is usually what an environmental health officer asks for on a commercial kitchen or a plant-heavy scheme.

Councils usually expect a suitably qualified author for acoustic work, and professional review is strongly advised. PlanScope's professional directory is one route to a reviewer for a draft.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.