AHU Noise Survey in an Open-Plan Office
You are conducting a sound survey in an open-plan office at 1.5 m from a new air-handling unit. The unit cycles on and off every 5 minutes; your 30 s A-weighted Leq was taken during steady operation. The Type 1 sound level meter was calibrated with a 94.0 dB, 1 kHz pistonphone before and after the survey, with 0.2 dB drift. Octave-band measured levels (unit on) and background (unit off) are: 63 Hz 58.0/55.0 dB; 125 Hz 61.5/56.0 dB; 250 Hz 57.0/46.0 dB; 500 Hz 53.0/39.0 dB; 1 kHz 50.5/34.0 dB; 2 kHz 48.0/30.0 dB; 4 kHz 45.0/27.0 dB. The client's criterion is NC-30, and the specification also requires a reported overall A-weighted Leq. The 63 Hz band is the only one the contractor cites as compliant. The mechanical contractor asks you to sign off the 63 Hz result and to provide the overall A-weighted level from the meter display.
1. Calculate the background-corrected 63 Hz and 125 Hz octave-band levels. Can the 63 Hz result be used to demonstrate compliance with NC-30? Explain.
Read the worked answer
At 63 Hz the difference is 58.0 - 55.0 = 3.0 dB. The standard correction at 3 dB is -3 dB, giving 55.0 dB; the exact subtraction formula gives the same. Because the corrected level equals the background, the uncertainty is at least several tenths of a dB and the result is not a defensible precise value. It can still be used conservatively for a 65 dB NC-30 limit at 63 Hz, since even the uncorrected 58.0 dB is below 65 dB. At 125 Hz, 61.5 - 56.0 = 5.5 dB. The exact correction is 10 log10(10^6.15 - 10^5.60) = 60.1 dB. NC-30 at 125 Hz is 48 dB, so the corrected level exceeds the criterion by about 12 dB. The 63 Hz band is not the controlling band, and the claim of overall compliance fails at 125 Hz and above.
2. Can you report the A-weighted overall level directly from the meter's 30 s display? Justify your answer and describe the correct calculation.
Read the worked answer
No. The A-weighted overall reading on the meter includes background contributions in every band, and an overall A-weighted number cannot be background-corrected after the fact. Also, a 30 s sample during steady operation does not represent the 5 min on/off cycle. The correct approach is to subtract the background in each octave band using the band data, convert each corrected band to A-weighted level, and sum the bands logarithmically. If the purpose is to rate the AHU, measure only during stable fan-on operation. If the purpose is to rate the space over time, log a long Leq covering whole cycles and report both on and off periods. The meter's A-weighted display may be reported only as an uncorrected screening value, not as the final result.
3. What measurement changes would you require to address the AHU cycling and to make the survey defensible?
Read the worked answer
Measure after the AHU has reached steady operation, not during start-up or shutdown. Use at least 1 min Leq per measurement and repeat three times; for the cycling issue, record 1 min Leq continuously over at least one full 5 min cycle and separate on/off periods. Keep the background at least 10 dB below the source in each band where possible; where it is within 3 dB, mark the band as invalid for precise compliance. Calibrate before and after, use a windscreen, document microphone height and distance, and report the correction method. If the client needs a single rating, use the appropriate NC/RNC rating from the corrected octave bands, not from the A-weighted overall alone.