What a noise survey report should contain

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.

Most people reading this have a noise survey report on their desk and no way to judge it.

That is an understandable position. OSHA's noise standard says what you must measure and what records you must keep, but it does not set a format for the report. Report quality therefore varies a great deal, and there is no published standard to check yours against.

The checklist below has 12 items. If your report is missing items 4, 7 or 11, it will not do what you need it to do when someone challenges it.

The 12 items

1. Who was monitored, and why those people

The report should give worker names or IDs, job titles, departments and shifts. Most of all, it should give the reasoning behind the similar exposure groups (workers who do the same job the same way) and say why the workers monitored represent those groups.

A report that lists results with no reasoning is not defensible, because nothing shows that the results describe anyone beyond the people who were monitored.

2. The shift length each average was calculated against

The report should not assume an "8-hour average" by default. Each time-weighted average (TWA) should be calculated against the actual shift length. If your plant runs 12-hour shifts and the report was calculated against 8, the numbers are wrong.

3. Full-shift personal dosimetry results for each worker

These are the results from the noise dosimeters the workers wore. For each worker, the report should give the 8-hour TWA, the dose percentage, the maximum level and the peak level. All four are needed, because a dose percentage alone does not let you check the math.

4. Dosimeter settings

Threshold, criterion level, exchange rate, weighting and response. The threshold is the level at which the dosimeter starts counting, and the criterion level is the 8-hour noise level that equals the full allowed dose. The exchange rate is the number of added decibels that cuts the allowed exposure time in half. For hearing conservation program work, the settings should read: 80 decibel threshold, 90 decibel criterion, 5 decibel exchange rate, A-weighting, slow response.

There is a second setup as well. Table G-16, OSHA's table of permissible noise exposures, sets the permissible exposure limit (PEL). That is the number that decides whether engineering noise controls are required.

OSHA's Technical Manual says to measure exposure for Table G-16 with a 90 decibel (dBA) threshold, so noise below 90 decibels is not counted in that average. The 80 decibel threshold is for deciding who belongs in the hearing conservation program.

These are two different measurements of the same worker, and modern dosimeters will log both at once. A thorough report says which setups were run and reports each result against the limit it belongs to.

The threshold is the setting that matters most. OSHA's noise standard requires the measurement to include all continuous, intermittent and impulsive sound from 80 to 130 decibels. A dosimeter run at a 90 decibel threshold discards everything between 80 and 90 decibels, which is the very range that decides whether a worker lands above or below the 85 decibel action level.

If the report doesn't state the threshold, you cannot tell whether the results understate what you must do.

5. Calibration records, before and after

The report should show the calibration check of each dosimeter against an acoustic calibrator, before it went out and after it came back, with the values.

The post-shift calibration is the one that gets skipped, and it's the one that matters. Without it you cannot show that the dosimeter didn't drift during the shift, and results from an instrument that drifted are not defensible.

Each dosimeter's make, model, serial number and last laboratory calibration date should also appear.

6. Area sound level readings, with locations

The readings should be tied to a floor plan or a documented grid, not listed as numbers with no positions. This is what you'll use to plan noise controls and decide where signs go, and it is also what allows a future survey to be compared against this one.

7. A comparison against both the action level and the PEL, not only the PEL

A report that only tells you who is over 90 decibels has answered the less important question. 85 decibels is the number that creates duties: enrolling workers in the hearing conservation program, audiometric testing, providing hearing protectors, training and recordkeeping. 90 decibels triggers the requirement for noise controls.

You need both lines drawn, clearly, for each worker.

8. A clear statement that results are not adjusted for hearing protection

Exposure for action level purposes is calculated without subtracting the attenuation of hearing protectors. OSHA's noise standard requires it to be computed "without regard to any attenuation provided by the use of personal protective equipment." The report should say so. If it is unclear whether attenuation was subtracted, the numbers may not mean what you think.

9. An evaluation of whether your hearing protectors are adequate

Does the protection in use now reduce the measured noise to the required level? The standard sets two targets: an 8-hour TWA of 90 decibels, and 85 decibels or below for any worker who has had a standard threshold shift (STS), a measured drop in hearing.

The better version of this evaluation does not rely on the Noise Reduction Rating (NRR) printed on the package. It uses octave band data, which are readings by frequency band, and the NIOSH long method.

This matters most where low-frequency noise dominates, because many protectors fall well short of their rated number there. → Octave band analysis

10. Field conditions and notes

The report should record what was running and what wasn't, the production rate compared to normal, whether noise controls were working, and anything out of the ordinary about the day.

This is what makes the report mean something in 18 months. A survey done during a slow week, with no note of it, will read as a clean result and will quietly misstate your plant.

11. What the results require, and by when

Results with no next steps are a lab report, not a survey. A useful report states plainly:

• Which workers belong in the hearing conservation program
• That baseline audiograms are due within 6 months of first exposure at or above the action level. Where mobile test vans are used, the deadline is 12 months, with hearing protection required after 6 months until the audiogram is done
• That workers exposed at or above 85 decibels must be told their results
• Whether feasible engineering or administrative controls are now required
• What would trigger repeat monitoring

12. Who prepared it, and the date

The report should say who did the survey, their qualifications, the date of the measurements and the date of the report. A report with no name on it is worth very little in a dispute.

Two things a good report also does

It states its limitations. It says which areas were not surveyed, which shifts were not covered, and which workers were not monitored and why. A report that claims to be complete when it is not is worse than one that is honest about its coverage, because you will rely on it for something it cannot support.

It compares against earlier data. If there are previous surveys, it says whether exposure has changed and what changed in the plant, because the trend often tells you more than the current snapshot.

How to check the report you already have

Ask five questions, in order of how often reports fail them:

1. Does it state the dosimeter's threshold setting? If not, you cannot trust its calls on who is over the action level. This is the most common omission, and the one that matters most.
2. Are there post-shift calibration records? If not, the results are not defensible.
3. Does it name who is over 85, not only over 90? If it only covers the PEL, it has answered the wrong question.
4. Does it explain the similar exposure groups? If it is only a list of people, it describes only those people.
5. Does it tell you what to do next? If not, you're holding data, not a survey.

If your current report fails two or more of these, it's worth resolving before your next inspection.

Audiometric testing, worker training and running the hearing conservation program are outside the four fields we work in. The noise survey and its report are our work.

→ Noise surveys · → See what our reports look like · Noise survey cost