Verifying a plant noise limit after installation

A plant noise verification condition is discharged by a measurement, not by a calculation, and the measurement has to be taken with the plant running in the state the condition describes at the position the condition specifies. Most failures at this stage are procedural rather than acoustic: the test was run in the wrong weather, at the wrong position, with the defrost cycle never observed, or against a limit the report quotes in different terms from the condition. Read the condition wording before you book the visit, and quote it verbatim in the report.

The usual form of words gives you a limit, a position, a method and a deadline. Each of those is a trap.

What the verification condition usually requires

A typical condition sets a rating level for noise emitted from plant, expressed in dB LAr,Tr, at a stated assessment position, determined in accordance with BS 4142:2014+A1:2019, with a verification report to be submitted to and approved in writing by the local planning authority within a stated period of first use.

Four variants cause most of the confusion.

An absolute limit in LAeq rather than a rating level, which removes the acoustic feature corrections from the comparison and makes the test easier. A limit expressed relative to the background sound level, such as 5 dB below the measured background, which obliges you to re-measure the background at the time of the test. A limit at a position described loosely, such as "the nearest residential boundary", which has to be pinned down with the environmental health officer before the visit. And a condition with no stated method, where BS 4142 is implied but not required.

Where the condition is relative to background, you are running two measurements rather than one, and the background may have changed since the original assessment was written. Re-measure it at the same time of day as the limit applies to, and expect the officer to compare your figure with the one in the original report.

Agree the position, the method and the operating states in writing with the officer before testing. A twenty minute email exchange prevents a second visit.

Measurement conditions needed for a valid test

Weather and instrumentation invalidate more verification reports than anything else.

Wind speed at the microphone has to be low enough that the measurement is not contaminated, with a windshield fitted and the speed recorded. No rain. No wet road surfaces where road traffic is part of the residual sound. And no extraneous events, which means a log of anything that occurred during the measurement rather than a silent average.

Instrumentation needs a class 1 sound level meter with a current laboratory calibration, field calibration before and after the measurement with the drift recorded, and the serial numbers of meter, calibrator and microphone in the report.

Residual sound is the measurement people skip. Measure with the plant off, at the same position, over a comparable period. The specific sound level is then obtained by logarithmic subtraction, and that subtraction becomes unreliable once the difference between total and residual falls below about 3 dB. Where that happens, say so and report the specific sound level as being at or below a stated figure rather than inventing a number.

Measurement period has to match the condition. If the limit is a night-time figure, the visit happens at night. A daytime test against a night limit is not a test.

Plant operating states to test

This is where a competent verification separates itself from a box-ticking one. Test the plant in every state it will actually occupy, rather than the state it happens to be in on the day.

1. All plant running simultaneously at maximum duty, which is the worst case the condition is normally written against.
2. Each item individually, with the others off, to identify the dominant source. Without this you cannot design remediation if the test fails.
3. Part-load and low-speed operation, because inverter-driven fans and compressors can be more tonal at intermediate speeds than at full speed.
4. Defrost cycle on any heat pump, chiller or refrigeration pack. This is frequently the loudest state, it lasts a few minutes, and a test run on a warm afternoon will never see it. Force it where the controls allow, and say in the report if you could not.
5. Night setback or reduced mode, where the condition sets a lower night limit and the plant is meant to drop into a quieter state. Verify that it does.
6. Extract systems at boost as well as at normal duty.
7. Any periodic event such as a sprinkler pump test or a generator exercise run.

Record which states were tested, which were not, and why. An honest gap is defensible. A silent one is not, and it is the first thing a reviewing officer looks for.

Dealing with a failed test

Do not submit a failing report on its own. Submit it with a remediation proposal, because a bare failure invites the authority to refuse to discharge the condition and leaves you with an unlawful development and no agreed way forward.

Work through four steps.

Identify the dominant source from the individual measurements and the octave or third-octave data. A 250 Hz peak points at a fan. A broadband hiss points at airflow. A discrete high-frequency tone points at a compressor or an inverter.

Quantify the shortfall in decibels at the assessment position, and translate it into the reduction required at source, allowing for distance and any screening already present.

Check whether an acoustic feature correction is driving the failure rather than the level. A 4 dB tonal correction on a plant item that is otherwise 2 dB inside the limit is a tonality problem, solved by changing the fan speed or fitting a different blade, not by adding mass.

Then choose the measure. Present the options and the predicted performance rather than a single answer, because the authority will usually accept a scheme that is justified over one that is asserted.

Remedial attenuation and retesting

Measures sit in a rough order of cost and effectiveness, and the first two are often ignored because they are unglamorous.

Reducing fan or compressor speed, or imposing a night setback through the controls, costs nothing in capital and something in performance. It is the fastest route to compliance and the easiest to undo later, which is why the authority will want the setting recorded and the control strategy described.

Relocating or reorienting the unit, so the discharge faces away from the receptor, can deliver a useful reduction for the cost of pipework.

A barrier or screen works only where it breaks the line of sight between source and receiver, needs adequate surface mass to avoid transmitting through, and must be sealed at the base. Expect a modest insertion loss rather than a transformation, because diffraction over the top limits what a barrier achieves.

An acoustic enclosure with attenuated ventilation openings delivers more, at the cost of airflow and therefore of plant efficiency, which the mechanical engineer has to sign off.

Attenuators and acoustic louvres on ducted systems are the predictable option, with manufacturer insertion loss data per octave band, and the pressure drop has to be checked against the fan curve.

Anti-vibration mounts and flexible duct connections address structure-borne transmission, which a BS 4142 airborne measurement will not diagnose. Where the complaint is a low-frequency hum inside a dwelling and the external measurement passes, suspect structure-borne transmission and investigate separately.

Retest in the same positions, by the same method, in the same operating states, and present before and after figures side by side with a statement of what changed.

Reporting format the authority expects

A verification report is a compliance document and it should read like one. Ten elements.

Instrumentation, serial numbers and calibration records. Measurement positions on a plan, with photographs showing the microphone in place and the distances to reflecting surfaces. Dates, times and durations. Weather including wind speed and direction. A plant inventory with makes, models and duties. The operating states tested, with those not tested and the reason. Raw time history and frequency data. Corrections applied, with justification for each. Residual and, where relevant, background levels. And a conclusion that quotes the condition verbatim and states compliance or otherwise against its exact wording.

Add an uncertainty statement. A report claiming compliance by 0.5 dB without acknowledging measurement uncertainty is weaker than one claiming compliance by 0.5 dB and explaining why that is still a sound conclusion.

The report belongs with the original Noise Impact Assessment and, where kitchen extract is involved, with the Ventilation and Extraction Statement, so the officer can see whether the installed plant matches what was assessed.

Record keeping for future complaints

Discharge of the condition is not the end of the matter, and the paperwork you keep decides how the next few years go.

Keep the verification report, the calibration certificates, the raw data files, the commissioning records and, above all, a written record of the control settings: fan speeds, timers, setback schedules and any limit programmed into the building management system. The first thing an environmental health officer asks after a complaint is whether the plant is still operating as tested.

Entertainment noise from a venue carries a parallel licensing regime, described in planning control and licensing control of venue noise, and a plant condition does not reach it.

Keep a maintenance log. Bearings, mounts and fan blades degrade, and a unit that passed at commissioning can drift. A documented maintenance regime is also the evidence that supports a best practicable means defence if a statutory nuisance case is ever brought, and that regime runs alongside rather than instead of the planning condition. Approval of a verification report does not prevent a nuisance action.

Where long-term monitoring is proposed instead of a spot test, the limits of unattended equipment matter, and they are set out in unattended noise monitoring: duration, position, validity.

Enter your postcode on PlanScope to see which noise and ventilation reports a site is likely to need.

This is general guidance, not legal advice; check the position with your local planning authority or a qualified professional before relying on it. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. For specialist technical reports, qualified professional input may be required. We do not guarantee planning approval.