A waste-transfer station report that subtracts its own penalties

The fatal error is the direction of the corrections. Under rating schemes of the ISO 1996-1 type, adjustments for tonal and impulsive character are added to the measured or predicted level because such noise is more annoying at equal LAeq; a clearly audible tone typically attracts about 5 dB, up to 6 dB. A fixed 10 dB 'intermittency allowance' subtracted in the source's favour has no technical basis - intermittency is already accounted for by using an energy-averaged Leq over the assessment period. Applying both in reverse has turned an exceedance into an apparent 7 dB margin.

On the report's own prediction, the level is 53 dB(A) against a 45 dB(A) criterion, an 8 dB exceedance, and 17 dB above the 36 dB(A) background, so masking is not available as an argument. Properly rated, with say +5 dB for the compactor and fan tones and +5 dB for impulsive bangs and reversing alarms, the assessed level is of the order of 63 dB(A) - 18 dB over the criterion.

The other assumptions cut both ways and should be identified as such: single downwind case is worst-case for propagation, hard ground is conservative, and omitting the bund is conservative. Against that, manufacturer sound power data is typically optimistic by 2-3 dB and should be specified with a margin. The net position is a substantial exceedance, not compliance.

I would require: octave-band sound power levels (63 Hz to 4 kHz) for the compactor, fans and vehicle fleet, from a certified laboratory or in-situ measurement, not a single A-weighted figure; vehicle pass-by source levels and a realistic night-time movement count; a receiver set that includes the care home at 60 m as well as the dwellings at 140 m, since the nearest receptor is not automatically the controlling one but here it is likely to be; a barrier calculation for the 6 m bund with the path difference and frequency-dependent insertion loss, because a bund that screens mid frequencies will do little at 63-125 Hz where compactor rumble sits; a meteorological treatment consistent with the chosen propagation standard, with the long-term average distinguished from the worst-case downwind case; and background data at the care home, not only at the dwelling.

The care home matters for a reason beyond proximity. Sleep disturbance and night-time awakening are the endpoints of concern, and the relevant criteria are usually expressed at the facade or inside the bedroom, with a lower limit at night. A care home population is more susceptible to awakening and less able to relocate bedrooms. Averaging the exposure over an evening of intermittent compactor runs and vehicle movements also hides the event structure: a small number of noisy events can dominate the sleep response even where the Leq looks moderate.

Recommended conditions, framed so they can be tested:

1. The compliance metric is the rated level, with character corrections applied as additions and the correction basis stated in the consent.
2. Night-time limits (23:00-07:00) at the care home and the nearest dwellings, in dB(A) and with a low-frequency descriptor, plus a limit on the number of night-time vehicle movements - preferably none.
3. Source controls as conditions, not aspirations: enclosure or acoustic hood on the compactor, low-noise broadband reversing alarms or a banksperson instead of reversing alarms at night, and silencers on fans.
4. A 3-5 dB design margin over the predicted level, so the prediction is not the design target.
5. A pre-commissioning and post-commissioning test with a defined method, positions, weather window, plant operating states, and a named independent reviewer; non-compliance triggers further mitigation, not renegotiation.
6. A noise management plan with a complaint line and log, and a review after twelve months of operating data.

I would also record in my advice that the original report's correction methodology should not be relied on for the determination, and give the authority the arithmetic showing why.