Rooftop plant noise assessed to BS 4142

A plant noise assessment roof scheme under BS 4142:2014+A1:2019 comes down to one comparison: the rating level of the plant at the nearest noise-sensitive receptor against the background sound level there, in the same period. Everything else in the report exists to make that comparison defensible. Where the rating level exceeds the background by around 10 dB, the standard indicates a significant adverse impact depending on context. Around +5 dB indicates an adverse effect, and a rating level at or below background is an indication of a low impact.

Establishing the sound power level of every item

The assessment is only as good as the source data, and this is where most reports fail before they get to the arithmetic.

Every item needs an octave band sound power level from the manufacturer, not an overall dB(A) figure from a sales sheet. Air source heat pumps, chillers, condensers, air handling units, kitchen extract fans, generators, lift motor rooms, dry air coolers and the fan on a packaged unit all belong on the schedule. So does anything that will be added later by a tenant fit-out, because that is what gets built.

Record for each item the make and model, the sound power level by octave band, the operating mode the data relates to, and whether the figure is measured or calculated. Watch three traps.

• Data quoted at part load or in a night-quiet mode that the building management system will not actually hold.
• Sound pressure at 1 m or 10 m quoted as though it were sound power, which understates the source by roughly 8 to 28 dB depending on the reference distance and radiation assumption.
• A single figure for a unit with several fans, where the real installation runs three of them.

Where manufacturer data is unavailable at design stage, say so and carry an allowance, then make the final selection subject to a condition. An assessment built on assumed data and presented as certain is worse than one that states its assumptions.

Cumulative operation and the worst case scenario

Councils assess the site, not the individual unit. So the specific sound level has to reflect everything running together in the period being assessed.

Set out an operating scenario table: which items run during the day, which run at night, and what the plant does on the hottest and coldest days of the year when everything is at full duty. Then assess the period that gives the highest rating level relative to background, which is almost always the night, because background falls faster than plant output does.

Logarithmic addition matters here. Two identical units together are 3 dB louder than one, four are 6 dB louder, and ten are 10 dB louder. A scheme that passes on one condenser and fails on four is common, and the arithmetic is quick enough that there is no excuse for discovering it at the condition discharge stage.

Also state the reference time intervals you used. BS 4142 works to a one hour reference time interval during the day, from 07:00 to 23:00, and 15 minutes at night, from 23:00 to 07:00. A report that assesses night plant over an hour has departed from the standard without saying so.

Reflection from parapets and adjacent facades

Roof plant sits in a reflective environment, and ignoring that is the second commonest technical error.

Three effects need to be in the calculation. A unit standing on a hard roof deck radiates hemispherically rather than spherically, which adds 3 dB compared with free-field radiation. A parapet or plant screen close behind a unit reflects sound outwards, and where the source sits in a corner between two surfaces the effect compounds. And a facade opposite, across a courtyard or a narrow street, returns sound towards receptors on the other side.

Parapets cut both ways. A parapet taller than the plant and close to it breaks the line of sight to a lower receptor and gives useful screening. The same parapet, lower than the plant, gives no screening and still reflects. Draw the sections before deciding which it is.

Courtyards deserve particular care. A lightwell surrounded by four hard facades with plant at the top behaves very differently from an open roof, and a simple point source calculation will underpredict.

Attenuators, enclosures and screens

Mitigation choices have to be stated by octave band insertion loss, because a claim of "10 dB attenuation" without a spectrum cannot be checked.

Inline attenuators on ducted systems are the most reliable option, with published insertion loss per band and a stated pressure drop. The pressure drop is the constraint: a fan resized to overcome it can end up louder than the original, which defeats the purpose.

Acoustic enclosures work for condensers and heat pumps, with two limits. They need free area for airflow, so the achievable insertion loss is usually modest, and a restricted airflow reduces the efficiency of the equipment inside. Louvred enclosures with lined plenums do better than sheet metal boxes.

Screens and barriers only work when they break the line of sight between the source and the receptor by a real margin, and their benefit falls away for low frequency content. A 2 m timber screen around a condenser at high level, with a receptor window on the sixth floor opposite, achieves close to nothing. The geometry decides the result, which is the same point made in acoustic fences and when a barrier actually reduces noise.

Anti-vibration mounts belong in the same section. Structure-borne transmission into a flat below a plant room is assessed differently from airborne sound, and a plant room above habitable rooms needs a resilient arrangement specified by an engineer rather than a proprietary mat named in a schedule. Where the sensitive room is directly beneath the source, the problem is closer to sound insulation between a venue and the flat above.

Night-time operation and setback

Night is where consents are won or lost, so design for it from the start.

Three moves reduce the night rating level at lower cost than mitigation added later. Set the plant back from the parapet edge nearest the sensitive receptor, so the parapet screens it and the distance helps. Locate items with tonal content, such as compressors and dry coolers, on the side of the roof away from residential frontages. And specify a genuine night setback mode, locked in the control strategy, with the reduced sound power level evidenced by the manufacturer.

Character corrections are applied to the specific sound level before comparison with background. Tonal content, impulsive content and intermittency each attract a correction under BS 4142, and a fan with an audible tone can add several decibels to the rating level without any change in the measured level. A 44 dB fan with a clear tone can be assessed as a 48 dB source. That is why tonality is a design question, not a reporting question.

State the background sound level honestly. A single 15 minute measurement on a wet Tuesday in a sheltered courtyard is not a representative night-time background, and an unrepresentatively high background inflates the headroom the scheme appears to have. Environmental health officers check this first.

Verification testing after installation

Expect a condition requiring post-installation verification, and write the assessment so it can be verified.

The condition normally requires measurement within a defined period after commissioning, at an agreed position, with the plant operating in its assessed mode, and a report demonstrating the approved limit is met. Make that achievable by stating in the application the measurement position, the mode, and the numeric limit you are accepting. Where the limit is expressed as a rating level not exceeding the background at the nearest receptor, say which receptor and which background figure.

Two practical points. Agree in advance what happens if the background at the time of testing is higher than at the assessment, because that can mask a non-compliant installation. And keep the commissioning data, because a later complaint will be investigated against the statutory nuisance provisions of the Environmental Protection Act 1990 rather than against your planning condition.

Drawings the assessment has to be read with

A Noise Impact Assessment for roof plant is unusable without drawings, and officers will ask for them.

Provide a roof plan showing every item numbered to match the source schedule, with screens and enclosures drawn to their real dimensions. Provide sections through the worst-case source and receptor pair, showing the line of sight and any screening. Provide an elevation showing screen heights against plant heights. And mark the receptor positions on a site plan with their floor level and window type, because a first-floor bedroom window and a sixth-floor bedroom window on the same facade give different results.

Where the plant serves a kitchen extract system, the acoustic case and the odour case have to be consistent with each other, since both are usually resolved by the same duct and discharge point. The Ventilation and Extraction Statement should reference the same schedule of equipment and the same discharge height rather than describing a different installation.

PlanScope drafts a Noise Impact Assessment structured to UK planning policy from your project details and your own measured or manufacturer data, with "TO BE CONFIRMED" flags where source levels or background figures are missing. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. For specialist technical reports, qualified professional input may be required, and councils expect a BS 4142 assessment to be signed by a suitably qualified acoustician.

This is general guidance, not legal advice. Check the position with your local planning authority or a qualified professional before relying on it. Enter your postcode on PlanScope to see which reports your scheme is likely to need.