Nighttime Warehouse Survey: Sampling Frame and Facade Correction
A logistics warehouse applies for an operating permit. The applicant submits a noise survey supporting compliance with a nighttime limit of 45 dBA Leq(8 hour) at residential property lines, night defined as 22:00–06:00. The survey covers three nights, Monday to Wednesday, at a single receptor. Each night a technician takes a 10-minute sample at the top of each hour using a Class 1 meter mounted on a balcony railing 3.2 m from the building facade and 1.8 m above the balcony floor. The three nightly Leq values are 46.2, 44.8 and 45.5 dBA. The report averages them arithmetically to 45.5 dBA and concludes compliance. No wind or precipitation data are recorded, although it rained during the second night. Truck movements at the loading dock, 90 m away, occur in bursts of 20–40 seconds, and the consultant notes that 'weekend operations are similar.' The complainant lives in a single-story house 15 m from the sampled building.
1. Evaluate the adequacy of this survey. Name the specific defects and state the direction of bias each introduces.
Read the worked answer
- Microphone position. 3.2 m from a facade, on a balcony railing, is closer than the ≥3.5 m practice; the alternative is a documented −3 dB facade correction. A railing above a balcony slab can read several dB high. Levels are biased upward - but the method is invalid regardless of direction.
- Arithmetic averaging. Nightly Leq values must be energy-averaged: 10·log10[(10^4.62 + 10^4.48 + 10^4.55)/3] ≈ 45.5 dBA. Same figure here, wrong procedure.
- Not an 8-hour Leq. Ten minutes out of each hour is 80 minutes of data for a 480-minute period. Truck pass-bys of 20–40 seconds fall between samples and are missed, biasing the night low.
- Sampling frame. Monday–Wednesday only; Friday and Saturday nights and seasonal peaks are unrepresented, so 'weekend operations are similar' is an assertion, not evidence.
- Receptor selection. The complainant's single-story house 15 m away has different geometry - ground effect and possible screening - from the sampled balcony.
- Weather. The rainy night is invalid; no wind data were gathered, and wind on a microphone creates pseudo-noise.
- No background or tonality data for the dock or refrigeration units.
Note also that the report's own conclusion contradicts its number: 45.5 dBA exceeds the 45 dBA limit.
2. Design a survey that would actually support a compliance determination.
Read the worked answer
Use continuous logging - 1-second or 1-minute Leq plus event records - for at least seven consecutive nights including Friday and Saturday, at the worst-case property-line point and at the complainant's facade.
- Microphone 1.2–1.5 m above ground, ≥3.5 m from walls, or with a documented correction; windscreen fitted.
- On-site weather logging; exclude periods of rain and wind above 5 m/s, and report how much data were excluded and why.
- Calibration check before and after each night with a traceable calibrator, recorded with instrument serial numbers.
- Background nights with the facility idle, to separate the source contribution.
- One-third-octave checks at the property line for tonality from the dock and refrigeration units.
- Operating assumptions documented: truck volumes, hours, yard equipment, refrigeration cycling.
Report the 8-hour energy-average Leq, L10/L50/L90, event counts and maximum pass-by levels, as a distribution across nights rather than a single mean, with a stated measurement uncertainty. If the result is within uncertainty of the limit, say so rather than declaring compliance.