How a noise survey works, step by step

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.

We come to your plant, put noise dosimeters on your workers for a full shift, and map your sound levels. Then we tell you exactly which workers your hearing conservation program has to cover.

Noise data goes stale quietly. Equipment changes, production moves or a line gets added, and the survey on file ends up describing a plant that no longer exists.

That matters more with noise than with most hazards, because a single result sets off a chain of obligations. One dosimetry result at 85 decibels (dBA) does not only mean "you have noise." It means you now owe audiometric testing, hearing protectors, annual worker training and recordkeeping, all of it with no end date, for every affected worker.

What we do

Personal noise dosimetry. Workers wear the dosimeters for the full shift, and this is the measurement that actually shows who is over the action level (explained below). → More on personal noise dosimetry

An area sound level survey and noise map. This shows where the noise is, how loud it is and how it travels through the building. It is the map you will use to decide where controls go and where hearing protection must be worn.

Octave band analysis. This breaks the noise down by frequency, from low to high, and you need that data before you can select an engineering control or choose hearing protection by the NIOSH long method. → More on octave band analysis

A review of noise controls. We look at what can realistically be done about the noise, what it would cost and what it would gain you.

When monitoring is required

OSHA's noise standard sets the trigger: when information indicates that any worker's exposure may equal or exceed an 8-hour time-weighted average (TWA) of 85 dBA, the employer must set up a monitoring program.

Read that trigger carefully, because it is lower than most people assume. The trigger is not "when you know workers are over 85" but when information indicates exposure may equal or exceed it. If your plant is loud enough that you are wondering, you are already inside the trigger.

The two numbers that matter:

Dose is the total noise a worker receives over the shift, expressed as a percentage of the daily amount OSHA allows.

Exchange rate: 5 decibels. Under OSHA's rate, every 5 decibel increase cuts the allowed exposure time in half. NIOSH recommends 3 decibels, which is more protective, and your survey should tell you about it even though it is not enforceable.

One detail that catches people. The standard says action level exposures must be determined without regard to any attenuation (noise reduction) from personal protective equipment. So you do not get to subtract your earplugs before deciding whether you owe a hearing conservation program. Handing out plugs does not put you below the action level; it only protects the worker inside a program you are still required to run.

What "loud enough to matter" looks like

As a rough guide from the field:

• You can't hold a normal conversation at 3 feet. You are likely in the mid-80s or above.
• Ears ring at the end of a shift. That is a temporary threshold shift, a short-term drop in hearing, and it is a warning sign, not a normal part of the job.
• Anyone has to shout to be understood. Assume you are at or over the action level until it is measured.

A rough guide is not a survey, however, and the reason is duration. Compare a stamping press at 100 dBA that runs 45 minutes a day with a compressor room at 88 dBA that someone sits in for 8 hours. The two produce very different doses, and the 88 dBA all day is the bigger problem. Only dosimetry can settle it.

How a survey runs, step by step

Step 1: A short planning call. We ask about your processes, headcount, shifts, layout and any earlier data, so we can work out how many similar exposure groups (SEGs) you have. An exposure group is a set of workers who do the same job the same way. As with air monitoring, the number of exposure groups sets the size of the job, not your headcount.

Step 2: A written scope of work and a fixed price. It gives the dosimeter count, the survey area, the deliverables and the turnaround time. → What drives the cost of a noise survey

Step 3: Measuring at your site. The method calls for checking each dosimeter against a sound calibrator before and after every use, and workers wear the dosimeters for the full shift. Area sound level readings go on a documented grid. Where engineering controls or hearing protector selection is in play, octave band analysis is added. Finally, the field notes record what was running, because a survey done on a slow day only describes a slow day.

Step 4: The report. It shows who is over the action level and who is over the PEL, and it includes your sound level map. A complete report also evaluates whether your hearing protectors provide enough attenuation for the noise measured, and it states what your results require you to do next. → What belongs in a noise survey report

Step 5: The follow-up conversation. If workers are over the action level, you have a program to build and clocks that start running. We talk that through with you, and the conversation is included.

Three details in how the standard reads

OSHA's noise standard has no ceiling limit. People often call 115 dBA OSHA's ceiling, but the standard has no such provision. That figure is simply the last row of Table G-16, OSHA's table of permissible noise exposures, where it is the "¼ hour or less" value. Treating it as a practical ceiling is reasonable, but calling it a stand-alone limit is not what the text supports.

The 140 decibel limit for impulsive noise is advisory, and it sits in a footnote. It appears in footnote 1 to Table G-16, which says exposure to impulsive or impact noise "should not exceed" 140 decibels peak. That is the wording of a recommendation, not of a mandatory limit. A different part of the standard says that if the peaks of a noise come 1 second apart or less, the noise is treated as continuous. The 140 decibel figure does not appear in that part.

Under OSHA, personal protective equipment can be how you meet the PEL. Above the Table G-16 levels, the standard requires feasible engineering or administrative controls. But the standard also says plainly what happens if those controls fail to bring levels within the table: personal protective equipment shall be provided and used to do it.

Note where the no-credit rule for hearing protection stops. The standard states it plainly as applying "for purposes of the hearing conservation program," and it does not carry over to meeting the Table G-16 levels.

Recordkeeping: two retention periods

OSHA's noise standard sets two very different record retention periods, and people often confuse them:

• Noise exposure measurement records: 2 years
• Audiometric test records: for as long as the worker is employed

The audiometric test record must include the worker's most recent noise exposure assessment. You therefore have to be able to retrieve your survey data well past its own 2-year minimum, because the audiometric file depends on it. A blanket 2-year purge of noise data will break the audiometric test records that refer to it.

Why combine a noise survey with air monitoring

If you owe both, do them on one visit, because the costly part of any survey is getting a qualified person to your plant. Workers can wear noise dosimeters at the same time as air sampling pumps, so it is the same shift, the same people and the same trip. Scheduling them separately means paying for travel twice and gaining nothing. → Air monitoring surveys