Noise dosimetry: how many employees do you need to sample?

This page explains what the rules and common practice call for. Every site is different, and what a survey or a filing needs at your facility is something we work out with you after we see it.

A compressor room at 88 decibels (dBA), occupied all day, produces a higher noise dose than a press at 100 dBA run for 45 minutes. That one fact determines how you sort workers into similar exposure groups (workers whose exposure over the shift is about the same) for a noise survey. It is also why sampling "the loudest machines" is the wrong plan.

The calculation is worth working through in full.

The worked comparison

OSHA's noise standard uses an exchange rate of 5 decibels and a permissible exposure limit (PEL) of 90 dBA as an 8-hour time-weighted average (TWA), so the allowed time is cut in half for every added 5 decibels:

T (hours) = 8 ÷ 2^((L − 90) ÷ 5)

Here T is the allowed time and L is the noise level in dBA.

A worker's dose is the time spent at a noise level, divided by the time allowed at that level.

Worker A is a compressor room attendant who spends 8 hours at 88 dBA. The allowed time at 88 dBA is 10.56 hours, which the table rounds to 10.6.

• Dose = 8 ÷ 10.56 = 75.8%

Worker B is a press operator who spends 45 minutes at 100 dBA, then 7.25 hours at 80 dBA somewhere else. Calculate each part, then add them.

• At the press: 0.75 ÷ 2 = 37.5%
• Somewhere else: 7.25 ÷ 32 = 22.7%
• Dose = 60.2%

A 50% dose defines the action level of 85 dBA, and workers at or above it must be in a hearing conservation program.

Both workers are over a 50% dose, so both are in the program. But Worker A, who never goes near anything above 88 dBA, carries the higher dose by a quarter.

Now change one variable. Suppose Worker B spends the other 7.25 hours at 75 dBA, not 80. The quiet hours now contribute nothing to the dose, for a reason the next section explains.

• Dose = 37.5% + 0 = 37.5%

That is below the action level, so Worker B is out of the program entirely.

The press and the 45 minutes are the same, yet the result under the standard is the opposite. What decided it was the quiet part of the shift, not the loud part.

The 80 decibel threshold decides everything

OSHA's noise standard requires the measurement to include all sound from 80 decibels to 130 decibels, whether it is continuous, intermittent or impulsive.

Note the bottom of that range, which is 80 and not 90.

A noise dosimeter set to a 90 dBA threshold, the level at which it starts counting, discards all the sound between 80 and 90 dBA. As the example above shows, that is the very range that determines who is above the action level and who is not.

The dosimeter will still produce a number, but that number will consistently run low, and it will be lowest for the workers whose dose comes from long stretches of moderate noise and not from short bursts.

The threshold is a setting the report has to state. A dosimeter set to an 80 dBA threshold supports a decision about the action level, and one set to 90 dBA does not.

A noise report should state the threshold, the exchange rate and the criterion level the dosimeter was run at. The criterion level is the 8-hour noise level that equals the full allowed dose. Confirm those three settings before the results decide who goes into the hearing conservation program.

→ Personal noise dosimetry · → How to read an industrial hygiene report

What splits an exposure group

An exposure group is defined by what a person's dose over the whole shift looks like, which depends on the following:

What does not split a group: the sound level printed on the equipment nameplate, the department or the job title.

What area monitoring can and cannot settle

OSHA's noise standard requires representative personal sampling, meaning dosimeters worn by workers who fairly represent the group. The requirement applies where any of three conditions makes area monitoring (readings at a fixed location) a poor fit:

• Workers move around, between areas with different sound levels.
• The sound level varies widely over the shift.
• A large part of the noise is impulse noise (sudden, sharp sound).

Read the closing words too. Personal sampling is required in those cases "unless the employer can show that area sampling produces equivalent results." So the standard does not flatly prohibit area data. It places the burden on the employer, who has to meet it with evidence and not by assertion.

Most manufacturing meets at least one of the three conditions, and often all three, which makes equivalence hard to prove.

Consider a worker who stays in one place, in steady broadband noise (noise spread across many frequencies) with no impulse noise. In that situation, area measurements plus a written time record can be equivalent, but where the worker moves between areas they will not be.

Whichever way that question falls, a sound level meter survey that maps the floor is still useful, because it helps you locate noise sources and plan engineering controls.

→ Area vs personal sampling in air monitoring · → Octave band analysis

So how many dosimeters?

OSHA's noise standard sets no headcount requirement. The number follows from how the exposure groups are defined:

The ranges assume two to three dosimeters for each group, which is enough to see whether the workers in a group really have about the same exposure.

If the first pass shows a wide spread within a group, that group was defined incorrectly and needs to be split. That is a normal and expected result of a first survey, not a failure of it.

Groups that sit close to the 50% dose line deserve more samples than groups that come in at 20% or 150%. Near that line, a wrong answer costs you in a way it does not anywhere else.

→ How to define similar exposure groups

When you may not need this

If a well-planned survey shows every group comfortably below a 50% dose, OSHA's noise standard sets no calendar for repeating it. The standard requires new monitoring whenever a change in production, process, equipment or controls increases noise exposure to the point that either:

• More workers may be exposed at or above the action level, or
• The attenuation (noise reduction) of the hearing protectors in use may no longer be adequate.

Note the direction: the trigger is an increase in exposure, tested against those two questions. It is not any change at all, and it is not the passing of 12 months. A change that lowers exposure, or leaves it where it was, does not by itself require a new survey.

An annual survey at a site with no such change is a schedule you have chosen, not an obligation the noise standard places on you. Keep the report, document the production conditions it was taken under, and measure again when a change meets that test.

→ Noise surveys · → Hearing conservation programs · → Standard threshold shift (STS)