Holiday lodges, Wales

The challenge

A rural landowner was resubmitting a scheme for holiday lodges after committee refused the first one. Noise was a stated reason: the units, the parking and the activity were held to disturb neighbours in a quiet spot.

The resubmission cut hard: half the units, capacity down by nearly two thirds, every unit solid-walled, parking reduced and moved, and nothing sited within thirty-five metres of a neighbouring house against a local guide figure of twenty-one. It still drew seven objections from neighbours and one from the community council, all about guests outdoors, hot tubs, cars coming and going, and, the recurring theme, scepticism that any policy about guest behaviour could be relied on.

There was also a technical obstacle. The sources in dispute were people talking and personal speakers, and the standard normally used for this kind of comparison explicitly does not apply to either. Using it anyway would have been a misapplication an objector could have taken apart.

What we did

We monitored for a full week at the corner of the nearest neighbour's garden closest to the lodges, so the measurement position represented both the person affected and the path the sound would travel. Despite the rural setting, a dual carriageway half a kilometre away and in clear view dominates the soundscape.

We assessed the change in the overall sound level rather than against an absolute limit, which is the appropriate framework for this kind of noise. Crucially, we used the quietest days recorded as the baseline rather than a typical one, because a quieter starting point makes any addition look bigger, which is the conservative direction.

Then we built a worst case that deliberately assumed the objectors were right. Every lodge occupied and generating noise at once, someone talking with a raised voice at every one, *and* a portable speaker playing at every one, which the site's own rules prohibit, all running continuously, with no credit for screening, ground or air absorption. We ran a second test for after eleven at night, dropping to normal conversation with the music stopped.