The 89.5 dB Plant That Believes It Is Under the Limit
An Abuja beverage plant runs two filling lines on a 10-hour shift, 06:00 to 16:30, with a 30-minute unpaid break and two 15-minute reliefs. Personal dosimetry on six filler operators gives 8-hour TWAs of 89.5 dB(A) (range 88.1 to 92.4), and a sound level meter reads 96 dB(A) at the operator position beside the capper. The plant issues foam earplugs labelled SNR 32 dB at the gate, with a poster. Three operators' annual audiograms show a 15 dB threshold shift at 4 kHz in the right ear; two of the three have no valid baseline. Audiometry is done in a room adjoining the compressor house where background noise measures 58 dB(A). Fourteen people have worked the line for more than eight years. Management's position is: 'The law is 90 dB, we are under it, and people wear plugs anyway.'
1. Test management's claim against the exposure data. Is the plant compliant? Show the arithmetic.
Read the worked answer
The claim fails on its own terms, twice over.
First, the exposure is not 8 hours, and the limit is an 8-hour limit. A 96 dB(A) level sustained for 10 hours normalises to an 8-hour TWA of 96 + 10 log(10/8) = approximately 97 dB(A). Compare that with the 89.5 dB(A) the dosimetry reported, and ask why they differ: the dosimeters almost certainly logged the relief periods and the break as quiet time, which is legitimate, but the operator position data and the personal data describe different things. The line is not 'under 90'.
Second, the dose calculation makes the margin obvious. Using a 3 dB exchange rate and an 85 dB(A) criterion, the permitted duration at 96 dB(A) is 8 / 2 to the power ((96-85)/3) = 8 / 12.7 = approximately 0.63 hours, about 38 minutes. A 10-hour shift at that level is a dose of roughly 1,590%. Even against a 90 dB(A) criterion the permitted duration is 8 / 2^2 = 2 hours.
Third, the legal argument is misstated. The 90 dB(A) figure is a limit on daily personal exposure over 8 hours, not on the level a meter reads at a machine, and it is not a threshold below which nothing is required. At 85 dB(A) a hearing conservation programme is triggered: assessment, signage, hearing protection provision and training, audiometric testing. At 89.5 dB(A) TWA the plant is in that territory on the personal data alone.
Fourth, the peak exposure. The capper generates impact noise that a TWA hides. Peak levels above 135 dB(C) matter independently of the average.
Conclusion to management. 'We are under 90' is wrong on the arithmetic and wrong on the law. The plant is a noise-controlled operation that has not been recognised as one.
2. Can the 4 kHz shifts be attributed to work? Assess the audiometric evidence and state what would make it defensible.
Read the worked answer
The pattern is suggestive but the evidence will not carry attribution as it stands.
What supports attribution. A 15 dB shift at 4 kHz is the classic noise notch, in three workers on the same line, in a plant measured at 96 dB(A) at the operator position, with fourteen people over eight years of exposure. That is a real cluster and not a coincidence.
What undermines it.
- The test environment is invalid. Audiometry in a room at 58 dB(A), adjoining a compressor house, produces thresholds that reflect the room. Test rooms are normally required to be in the region of 25-30 dB(A) or better; 58 dB(A) makes low-frequency thresholds unreliable and contaminates the very comparisons being made.
- Two of three have no valid baseline. Without a baseline you cannot demonstrate a shift; you can only demonstrate a current threshold, which may be pre-existing. Occupational attribution requires a change over time or a pattern consistent with exposure history.
- No clinical workup. No otoscopy, no history of ototoxic medication, firearms, music or previous employment, no age correction, no examination by an audiologist.
What I would do. Refer all fourteen long-service operators to an audiologist; retest after at least 14 hours without occupational noise; obtain or reconstruct baselines from pre-employment records; take occupational and non-occupational histories; apply age correction; and repeat in 30 days to establish reliability.
How the finding is worded. 'Three cases of suspected occupational noise-induced hearing loss, pending confirmation' with the evidential gaps set out. Writing 'confirmed NIHL' on this data would not survive a claims hearing, and would hand the employer a reason to dismiss the finding entirely.
3. Propose a control strategy with a defensible timeline, ranked by effectiveness, and say what will not fix this.
Read the worked answer
Ranked by reliability, not by ease.
1. Elimination and substitution, at the source. Is the capper needed at that speed, that duty cycle, that location? Line speed reduction, different closure design, or relocating the capper into a separate room are the only measures that actually reduce the energy entering the operator's ear.
2. Engineering controls. Acoustic enclosure around the capper with the feed and discharge apertures baffled; silencers on air exhausts; anti-vibration mounts; damped guarding panels; preventive maintenance, because a worn bearing turns a 96 dB machine into a 102 dB machine quietly and nobody notices except the operators.
3. Administrative controls. Remote monitoring and CCTV so the operator is not standing at the noisiest point; job rotation designed to keep individual daily exposure below 85 dB(A); clearly marked hearing protection zones at the line entry points; quiet refuges for breaks. Rotation must be engineered against dose data, not rosters.
4. Hearing protection, last and least reliable. Foam plugs at the gate with a poster is not a programme. It needs individual fitting, demonstration of correct roll-and-insert, field attenuation testing, a choice of at least two types, replacement on a schedule, and supervision. Even then, real-world attenuation is often 5-10 dB rather than the 32 dB on the packet, because the SNR figure is a laboratory label value, not a guarantee of fit.
Timeline. Immediately: dosimetry mapping by task, signage, mandatory HPD with fitting and training, audiometric referrals. Three months: engineering assessment and quote. Twelve months: enclosure and maintenance regime in place, with post-modification dosimetry to prove it worked.
What will not fix this. The poster. The plugs alone. A toolbox talk. Buying a quieter machine without checking its installed level. And an audiogram programme run in a room next to a compressor.