Noise surveys: measure first, decide with data.

A noise survey is not one number. The logger records a set of statistical indices every period, and each answers a different question. The LAeq is the equivalent continuous level, the energy average, and it is what BS 8233 internal targets and most planning conditions are written against. The LA90 is the level exceeded for 90% of the time, which is taken as the background sound level and is what BS 4142 compares plant against. The LAmax captures individual events, and it is what governs sleep disturbance where a railway or a delivery yard is the concern.

Reporting only the LAeq is the most common shortcoming we see in surveys carried out by others. A site can have a perfectly ordinary average level and a background low enough that any plant proposal fails, or an unremarkable average punctuated by pass-by events well above the threshold at which people wake. The indices have to be reported together, because the conclusion depends on which one governs.

Alongside the numbers, a survey should record what was actually audible. Where a level is driven by a specific source, an assessment that identifies it is worth considerably more than one reporting an anonymous figure, because mitigation depends on knowing what is making the noise.

The single biggest determinant of whether a survey survives scrutiny is whether it was carried out over a representative period. A weekday survey on an industrial estate misses the Sunday morning conditions that generate complaints. A survey in the school holidays misses term-time traffic. A survey in still weather misses the wind-induced levels that mask plant for half the year.

For most planning applications that means at least 24 hours of unattended logging, extended where the environment varies. Where a night-time background is the controlling figure, we would normally log several consecutive nights, because a single night can be unrepresentatively quiet or unrepresentatively noisy and there is no way to tell from one night's data which you have.

Weather matters more than people expect, and not just for equipment protection. Measurements taken during rain or in wind above roughly 5 m/s are unreliable, and a survey that does not record the weather conditions alongside the levels cannot demonstrate that they were valid. We log weather throughout and exclude affected periods rather than quietly averaging them in.

The fee is fixed against a defined scope: the number of measurement positions, the duration of logging, whether attended measurement is needed alongside the unattended logger, and which standards the assessment is written to. Those four variables account for almost all the difference between a straightforward survey and an expensive one.

What does not change the fee is travel. Most surveys are unattended, so equipment is deployed and collected, and distance affects our costs rather than yours. Nor does the result: we are paid to measure and assess, not to reach a particular conclusion, and a survey that finds a problem takes the same effort as one that does not.

What can change it is a change of scope after the fact. If the council asks for something the original condition did not require, or the scheme changes so that a different receptor becomes the worst affected, that is additional work and we will quote it before doing it rather than after. The one thing we will not do is discover extra cost once the survey is already on site.