Beryllium: any exposure at all triggers an assessment

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 of OSHA's health standards start counting at the action level, which is set below the permissible exposure limit (PEL). The beryllium standard does not.

OSHA's beryllium standard requires an exposure assessment for each worker who is, or may reasonably be expected to be, exposed to airborne beryllium. Any amount counts. The trigger is not “at or above the action level,” and it is not “at or above the PEL.”

The trigger is exposure itself, which is broader than the trigger in the lead, silica or cadmium standards. The hexavalent chromium standard is built the same way and also sets no action level threshold for initial monitoring.

That is why beryllium surveys get planned wrong. The action level is 0.1 micrograms per cubic meter of air (µg/m³), and it is tempting to read that number and assume it works like lead or cadmium. People then conclude they owe nothing until a sample shows a number. The obligation runs the other way: you monitor to find out where you stand, because the duty to find out comes first.

Where we find it

• Machining, grinding and deburring copper-beryllium alloy. Parts include connectors, bushings, non-sparking tools, injection mold inserts and oil-field components. The alloy is a few percent beryllium, and nobody on the floor thinks of it as a beryllium job.
• Dental laboratories. Grinding and polishing casting alloys that contain beryllium.
• Aerospace. Beryllium and beryllium-aluminum parts, optical structures, and brake and structural parts.
• Abrasive blasting with coal slag. Coal slag carries a trace of beryllium, and this source surprises people most, because the operator bought an abrasive, not a beryllium product.
• Foundries and smelters that melt metal containing beryllium.
• Scrap and metal recycling. Torch cutting and shredding mixed alloys.

Exposure limits

The table shows OSHA's PEL, its short-term exposure limit (STEL) and the action level, followed by NIOSH's recommended exposure limit (REL). The PEL and the action level are 8-hour time-weighted averages (TWA).

Note the units. These limits are tenths of a microgram per cubic meter, which is 25 times below the hexavalent chromium action level and tens of thousands of times below a limit for nuisance dust.

At concentrations this low, the smallest amount the lab can measure is close to the amount you are trying to measure. Sample volume stops being a detail and becomes the whole plan.

Note also that the 2.0 µg/m³ value is a 15-minute STEL. It is not a ceiling limit, which is a level that must not be exceeded at any time. You show you meet it with a separate 15-minute sample taken during the task with the highest exposure, so someone has to know when that task happens.

How we sample for it

Two field errors account for most bad beryllium data.

The first is sampling for too short a time. A full-shift sample at 2 L/min collects enough dust on the filter for the lab to report a reliable result below 0.1 µg/m³. If the run is cut to three hours because the department shut down early, the lab may report “none detected” with a smallest measurable amount that is above the action level. That result does not say you are below the action level. It says the survey could not tell.

The second is contamination. At concentrations this low, gloves that touched copper-beryllium chips can contaminate a cassette. The result then comes back positive for a reason that has nothing to do with the air the worker breathed.

What your results require you to do

Once the initial exposure assessment is done, the results determine how often you repeat the monitoring:

• At or above the action level, but at or below the 8-hour PEL: periodic monitoring every six months.
• Above the 8-hour PEL: every three months.
• Above the STEL: every three months.

Discontinuing monitoring runs on two separate clocks, and each one needs two consecutive measurements, taken at least seven days apart:

• The 8-hour TWA. Periodic monitoring for the 8-hour TWA may stop when two consecutive measurements, seven or more days apart, are below the action level.
• The STEL. Periodic short-term monitoring may stop when two consecutive measurements, seven or more days apart, are below the STEL.

People sometimes cite a different paragraph of the standard for this (paragraph (d)(3)(iv)), but that paragraph covers initial monitoring, not the end of periodic monitoring.

Because the clocks are separate, clearing one does not clear the other. A grinding job that averages under 0.1 µg/m³ across the shift but spikes above 2.0 µg/m³ during wheel dressing keeps its obligation to monitor against the STEL.

The short-term sample is not an optional extra on a beryllium survey. It is half of what it takes to stop monitoring, and it needs two rounds a week apart, just as the 8-hour TWA does.

→ How often OSHA requires air monitoring, by substance

Beryllium usually arrives inside something else

Few plants buy beryllium. They buy copper alloy stock, dental casting alloy, coal slag abrasive or a container of mixed scrap, and the beryllium arrives inside the purchase. The safety data sheet lists it as a small percentage, and then machining or blasting releases it as dust fine enough to reach deep into the lungs.

Chronic beryllium disease is a sensitization response, meaning the body's immune system starts to react to beryllium. Once a person is sensitized, dose and disease stop following the steady pattern the rest of your program is built on.

That is why the standard's trigger for an exposure assessment is written the way it is. It is also why we will not agree in advance that a machining cell must be clean. We would rather sample it and tell you it is.

→ What an air monitoring survey costs in 2026