Industrial Noise Control — Enclosures, Barriers & Worker Protection | AcousPlan

The Noise Control Hierarchy

Industrial noise control follows a fundamental hierarchy that mirrors occupational health practice: eliminate or reduce the hazard at source; if that is not feasible, control the transmission path; if that is not feasible, protect the receiver. Each step down the hierarchy is more expensive per dB of reduction and less effective at protecting all workers simultaneously. The hierarchy is enshrined in UK law under the Control of Noise at Work Regulations 2005, Regulation 6, which requires employers to reduce noise exposure by means other than hearing protection wherever reasonably practicable.

This guide works through the hierarchy systematically, with calculations and worked examples for the most common industrial noise control interventions.

Regulatory Framework

UK: Control of Noise at Work Regulations 2005

The Control of Noise at Work Regulations 2005 (CNAW) implement the EU Physical Agents (Noise) Directive and establish:

The CNAW also requires that employers eliminate noise at source or reduce it to as low a level as is reasonably practicable, using the engineering controls described in this guide, before resorting to hearing protection as the primary control.

US: OSHA 29 CFR 1910.95

OSHA's Occupational Noise Exposure Standard requires:

• Feasible administrative or engineering controls when noise exceeds 90 dBA TWA (8-hour)
• Hearing Conservation Programme (monitoring, audiometric testing, HPDs, training) when noise exceeds 85 dBA TWA
• Maximum permissible exposure level: 90 dBA TWA (8-hour), 95 dBA TWA (4-hour), up to 115 dBA (15 minutes maximum)
• Dual hearing protection (muffs + plugs) where engineering controls cannot reduce below 100 dBA

Source Treatment

Source treatment is the most cost-effective intervention in the long run. It reduces noise at the point of generation, benefiting all workers in the facility simultaneously, with no need for hearing protection, no noise zone demarcation, and no management overhead.

Machine Isolation

Mechanical plant generates noise through two pathways: airborne radiation from vibrating surfaces, and structure-borne transmission through fixings and foundations. Isolation addresses the structure-borne path.

Anti-vibration mounts (AVMs): Rubber or spring isolators installed between the machine base and its supporting structure. Selection is based on the forcing frequency (typically the fundamental frequency of the rotating equipment) and the static load.

The natural frequency of the isolation system must be significantly below the forcing frequency for effective isolation. A rule of thumb: the natural frequency of the isolation system should be one-quarter to one-fifth of the forcing frequency.

For a compressor running at 1,500 rpm (25 Hz forcing frequency):

• Target isolator natural frequency: 25/5 = 5 Hz
• This corresponds to a static deflection of approximately 10 mm (δ_st = g/(4π²fn²) = 9.8/(4π² × 25) ≈ 0.01 m = 10 mm)
• Rubber mounts with 10–15 mm deflection are standard for this application

Low-Noise Equipment Selection

Specifying quieter equipment at the procurement stage is the most effective and lowest-lifetime-cost noise control measure. A 10 dB reduction in sound power at source eliminates the need for approximately £50,000–£200,000 of downstream noise control engineering for a typical industrial installation.

When comparing equipment options, require sound power level (LW, in dB re 10⁻¹² W) per ISO 3744 or ISO 9614, not sound pressure level at a single measurement point. Sound power level is an intrinsic property of the machine, independent of the acoustic environment; sound pressure level at a measurement point is useless for comparison across installations.

Path Treatment: Barriers

Barriers (acoustic screens, bunds, noise walls) intercept sound propagating from source to receiver. They work by blocking the direct path and causing diffraction at the top edge — reducing but not eliminating sound in the "shadow zone" behind the barrier.

Maekawa's Formula

The classical formula for barrier insertion loss (Maekawa, 1968):

IL = 10 × log10(3 + 20N)

Where N is the Fresnel number:

N = 2δ/λ = 2δf/c

• δ = path length difference (m): distance over the barrier minus the direct path distance
• λ = wavelength (m) = c/f
• f = frequency (Hz)
• c = speed of sound (343 m/s at 20°C)

Source to barrier to receiver geometry:

• Source height: 1 m
• Receiver height: 1.5 m
• Source-to-barrier distance: 10 m
• Barrier-to-receiver distance: 20 m
• Barrier height: 4 m

δ = 30.60 − 30.0 = 0.60 m

At 500 Hz (λ = 0.686 m): N = 2 × 0.60 / 0.686 = 1.75 IL (500 Hz) = 10 × log10(3 + 20 × 1.75) = 10 × log10(38) = 15.8 dB

At 125 Hz (λ = 2.74 m): N = 2 × 0.60 / 2.74 = 0.44 IL (125 Hz) = 10 × log10(3 + 20 × 0.44) = 10 × log10(11.8) = 10.7 dB

Barriers provide significantly less protection at low frequencies — a critical limitation for industrial sources with significant low-frequency content (compressors, generators, presses).

ISO 9613-2 Method

ISO 9613-2:1996 (Attenuation of Sound During Propagation Outdoors) provides a more complete calculation that accounts for:

• Ground reflection (constructive or destructive interference from ground-reflected ray)
• Atmospheric absorption (frequency-dependent, significant over long distances)
• Multiple diffractions (e.g., over a wide barrier)
• Screening by hills and embankments

Practical Barrier Design

For effective barrier performance:

• Surface mass: ≥ 10 kg/m² to avoid transmission through the barrier material dominating over diffraction. A 50 mm concrete panel (120 kg/m²) or 2 mm steel sheet (16 kg/m²) are both adequate.
• Height: Every 1 m of additional height above the direct line of sight adds approximately 1.5–2 dB of insertion loss. Doubling the barrier height does not double the insertion loss.
• Length: The barrier must be long enough that lateral diffraction around the barrier ends does not dominate. The barrier end should extend at least 4× the direct path distance beyond the source-receiver lateral axis.
• Surface treatment: Hard, reflective barriers can cause problems if reflections from the barrier face create hot spots. For large barriers adjacent to noise-sensitive receptors on the far side, absorptive cladding on the barrier face (NRC 0.70–0.90) reduces reflections.

Path Treatment: Enclosures

An enclosure is the most effective noise control intervention per dB of reduction, but also the most expensive and operationally challenging. Full enclosure of a machine can achieve 25–40 dB reduction; partial enclosure (acoustic hood, acoustic cabin) achieves 10–20 dB.

Enclosure Insertion Loss

The insertion loss of a noise enclosure is limited by two factors:

1. Transmission Loss (TL) of the walls The enclosure walls must have sufficient mass to prevent sound from transmitting through them. The mass law approximation:

TL ≈ 20 × log10(m × f) − 47 (dB) (for a single-leaf panel in the mass-controlled region)

Where m = surface mass (kg/m²) and f = frequency (Hz).

For a 2 mm steel panel (m = 16 kg/m²) at 500 Hz: TL ≈ 20 × log10(16 × 500) − 47 = 20 × log10(8000) − 47 = 78 − 47 = 31 dB

2. Reverberation inside the enclosure An untreated enclosure builds up a reverberant field inside, increasing the sound level inside by 6–10 dB above the free-field level. This reduces the effective insertion loss accordingly. Adding absorptive material to the interior walls (NRC 0.80+) reduces reverberation and improves insertion loss by the same 6–10 dB.

The combined insertion loss of an enclosure:

IL_enclosure = TL_wall + 10 × log10(α × S_interior / S_exterior)

Where α is the average absorption coefficient inside the enclosure and S is the respective surface areas.

For a machine inside an enclosure with 50% of interior surfaces treated (NRC 0.85):

• TL_wall = 30 dB at 500 Hz
• Mean interior α = 0.5 × 0.85 + 0.5 × 0.05 = 0.45
• S_interior ≈ S_exterior (enclosure close-fitting)
• IL = 30 + 10 × log10(0.45) = 30 − 3.5 = 26.5 dB

Enclosure Design: Practical Requirements

Access and maintenance: Enclosures must allow access for routine maintenance. Specify acoustic doors with seals (Rw 25–35) and acoustic access panels. Every unsealed penetration reduces insertion loss dramatically — a 1% open area reduces the maximum achievable TL to approximately 20 dB regardless of wall construction.

Ventilation: Machines inside enclosures require ventilation to remove heat. Ventilation openings must be fitted with acoustic splitter silencers. A standard 300 × 300 mm duct opening provides negligible attenuation; a 1 m silencer with 50 mm absorptive lining gives approximately 20–25 dB attenuation at 500 Hz.

Pipe and cable penetrations: All penetrations through enclosure walls must be sealed with acoustic sealant or lagged with mass-loaded vinyl. Unsealed conduit penetrations are a common failure mode.

Receiver Protection: Hearing Protection Devices

Hearing Protection Devices (HPDs) are the last resort in the control hierarchy — they protect only the individual wearing them, require correct selection and fitting, degrade over time, and impose communication difficulties. Despite these limitations, they are necessary where engineering controls cannot reduce exposure below action values.

HPD Selection

HPDs are selected based on their Single Number Rating (SNR, Europe) or Noise Reduction Rating (NRR, US) and their suitability for the noise spectrum.

The SNR method (Europe, EN ISO 4869-2): Assumed protection = SNR − 4 dB (the 4 dB penalty accounts for real-world fit variation) Effective noise exposure = Measured level − Assumed protection

For a wearer exposed to 95 dB(A) with earmuffs (SNR 28): Assumed protection = 28 − 4 = 24 dB Effective exposure = 95 − 24 = 71 dB(A) — well below the 80 dB action value.

The H/M/L method uses octave band levels and provides more accurate protection estimates where the noise spectrum is known. Prefer H/M/L for precision; use SNR for quick assessment.

Dual Protection

Where a single HPD is insufficient to reduce exposure below the action values, HSE guidance requires dual protection: earmuffs worn over earplugs. Dual protection does not add the individual SNR values — the combined attenuation is approximately 3–5 dB above the higher of the two devices (the limiting factor is bone conduction).

Hearing Conservation Programme

Any workplace where workers are exposed above 80 dB(A) LEP,d must have a Hearing Conservation Programme (HCP) including:

• Noise monitoring: Identify which workers are exposed above action values (initial survey + after changes)
• Audiometric testing: Baseline audiogram at employment start, annual testing thereafter for exposed workers
• Hearing protection provision: Appropriate HPDs available at all noise zones, replacement on request
• Training: Workers informed of noise risks, correct use of HPDs, programme administration
• Record keeping: Audiograms retained for 40 years per UK requirements