The Two Numbers

In California, Cal/OSHA’s noise standard lives at Title 8, Section 5097 of the California Code of Regulations. The federal equivalent is 29 CFR 1910.95. Both define two thresholds for an 8-hour time-weighted average (TWA) of employee noise exposure.

• 85 dBA TWA is the action level. When an employee’s 8-hour TWA reaches or exceeds 85 dBA, the employer must put a hearing conservation program in place for that employee. That means a documented program, audiometric testing, training, hearing protection availability, and recordkeeping. The action level is what triggers the program. It does not, by itself, mean an exposure is unlawful.
• 90 dBA TWA is the permissible exposure limit (PEL). When exposure reaches 90 dBA TWA, the employer is required to reduce that exposure through engineering controls or administrative controls. Hearing protection is a backstop, not a substitute. If controls aren’t feasible or can’t bring exposure below 90 dBA, hearing protection must reduce the exposure the employee actually receives.

A facility can be over the action level (85) and under the PEL (90). That’s a common position, and it’s the one that puts you squarely inside hearing conservation obligations without yet requiring engineering control work.

The Dose Concept

OSHA expresses noise exposure as a percentage of an 8-hour dose. The 90 dBA PEL is defined as 100% dose: an employee exposed at exactly 90 dBA for a full 8-hour shift has taken on 100% of the allowable daily exposure.

The 85 dBA action level corresponds to 50% dose. Half the PEL exposure, expressed in dB, is 5 dB lower, not half the number. This is because OSHA uses a 5 dB doubling rate (also called the exchange rate). Every 5 dB of additional exposure doubles the dose. A shift at 95 dBA is 200% dose. A shift at 100 dBA is 400% dose.

That doubling-rate framing matters because noise exposure is rarely flat across a shift. A welder spending two hours at 100 dBA and six hours at 75 dBA may end up with a TWA in the mid-80s, depending on the math. The decision about whether that employee belongs in the program isn’t made by glancing at peak readings. It’s made by a personal dosimeter that integrates exposure across the shift.

OSHA Noise Exposure Reference Chart

The 80 dBA row is worth noting. It isn’t a regulatory threshold for the employer, but it is the threshold most dosimeters are set to. Cal/OSHA’s measurement parameters track noise from 80 dBA upward, A-weighted, slow response, with a 90 dBA criterion and a 5 dB doubling rate. If a survey is done with different settings, the results aren’t comparable to the regulatory limits.

The Extended-Shift Wrinkle

The 85 and 90 dBA numbers assume an 8-hour shift. Many of the facilities we walk run longer shifts: 10-hour shifts in metal fabrication, 12-hour shifts in food and chemical processing, occasionally longer with overtime. When the shift is longer than 8 hours, the action level and PEL adjust downward, because the employee is being exposed for more hours.

Using OSHA’s 5 dB exchange rate, the adjusted action level works out to roughly 83.4 dBA TWA for a 10-hour shift and roughly 82.0 dBA TWA for a 12-hour shift. The PEL adjusts the same way, sitting about 5 dB above each number.

This catches facilities that look at an 85 dBA TWA from a 10-hour shift and assume they’re at the line. They aren’t. They’re already over the adjusted action level. Whenever we run a noise survey at a facility on extended shifts, the math is done against the shift length the employee actually works, not against a default 8-hour reference. Inspectors do the same.

Why the Survey Determines Enrollment, Not the Job Title

A common assumption is that “the loud roles” need the program and the rest don’t. In practice, enrollment is determined employee by employee, by measured exposure, using a sampling strategy that represents each job and area. Two employees with the same title can have different exposures if their tasks differ across the shift. An office worker who walks the production floor for an hour a day may sit far below the action level even though they spend time in a loud space.

The sampling strategy is part of the program. Section 5097 requires three things of the employer: identify employees exposed at or above the action level, repeat monitoring when production processes or controls change in ways that may affect exposure, and notify affected employees of their results. The list of who’s in the program isn’t static. When we work with a facility whose layout or equipment has shifted since the last survey, re-monitoring almost always lands on the project plan.

How This Connects to a Hearing Conservation Program

Crossing the action level isn’t a one-time event. It’s the start of an ongoing program that includes baseline and annual audiometric testing, annual training, hearing protector selection and fit, recordkeeping (noise measurements retained at least 2 years, audiograms retained for the duration of employment under Title 8 §5100), and program administration.

For a complete overview of what a compliant program contains and how facilities typically stand one up in California, see our hearing conservation program requirements guide. If you haven’t yet had noise monitoring done, that’s the diagnostic step that produces the TWA results discussed above. We cover the field-measurement side under sound monitoring.

Two related pieces in this cluster go deeper on adjacent questions. The audiometric testing piece covers what happens once an employee is enrolled: audiometric testing for hearing conservation walks through the baseline, the annual cycle, and standard threshold shift mechanics. For California-specific program requirements under Title 8 §5097, see Cal/OSHA hearing conservation program requirements.

What to Do With Your Numbers

If you have a recent noise survey, three checks usually resolve the immediate question:

If any of those answers raise a flag, the next step is either a re-survey or a closer look at the existing data with someone who can interpret it. We do both, and we build the program around the result.