Plant noise assessment: quiet plant, approved.

Manufacturer sound data is the usual culprit. Published figures are measured under defined conditions that rarely match the installation: a unit tested free-field behaves differently mounted in a corner between two hard walls, ducted plant loses or gains depending on the run, and multiple identical units at similar distances combine in a way that a single-unit figure does not capture.

The second recurring problem is the receptor. The nearest dwelling on a site plan is often not the worst affected. A first-floor bedroom window with a clear line of sight to a roof-mounted unit can be considerably more exposed than a ground-floor room a few metres closer but screened by a boundary wall. The assessment position has to be chosen from the geometry, not from the plan.

Third is operating pattern. Plant assessed at typical load understates the case where the controlling condition is peak demand on a still summer night, which is exactly when windows are open and background sound is lowest. The assessment should be made against the worst credible operating condition, not the average one.

Mitigation has a clear order of preference, and it runs from free to expensive. Moving the plant is first: distance, orientation and screening by the building itself cost nothing while the drawing is open. Selecting a quieter unit is second, and frequently the difference between models is larger than anything an attenuator will achieve. Only then do acoustic enclosures, attenuators and barriers become the answer.

Barriers are widely misunderstood. A barrier works by breaking the line of sight, and its benefit depends on how far the sound has to bend around it rather than on its mass. A barrier that is tall enough but has a gap at the base, or does not extend far enough sideways, will underperform its specification substantially. It also has to be sufficiently heavy to stop transmission straight through, which for most plant means a solid construction rather than a timber screen.

Whatever is specified needs to be verifiable. We state the performance the mitigation has to achieve, so that what gets installed can be checked against it and, where a condition requires post-installation verification, measured and signed off without argument.

Not all plant assessment is for planning. A significant part of this work is for operators whose equipment is already installed and running, and where a neighbour has complained or the council has made contact. The technical question is the same, but the context is not: the plant exists, its noise can be measured directly rather than predicted, and the outcome has consequences beyond a planning decision.

Measuring installed plant is more reliable than predicting proposed plant, because the uncertainty in manufacturer data and installation effects disappears. What it needs is care in separating the plant from everything else: measurements with the plant running and shut down, so the specific sound level and the background can both be established at the same position on the same night.

Where the assessment shows a genuine problem, saying so early is the cheaper path. A remediation designed and delivered voluntarily is almost always less expensive and less disruptive than one imposed under an abatement notice with a deadline attached.