Hearing protectors: types, ratings and derating

Hearing protectors are the last line of defense against noise. They are cheap, portable and, when properly selected, fitted and worn, highly effective. They are also the control most likely to fail without anyone noticing: a foam earplug pushed only halfway in, an earmuff seal broken by safety glasses, a protector removed for a few minutes in the loudest part of the shift. This lesson covers the main types of protectors, how they are rated, why laboratory ratings overstate real-world protection and how to estimate the protection a worker actually receives. Lesson 9 covers fitting, fit testing and day-to-day use.

HSE-104 Personal Protective Equipment, lesson 5, introduced protector types and ratings. This lesson goes further into the calculations and selection decisions that a program coordinator or supervisor needs.

Types of hearing protectors

No single type is best for everyone. Ear canal shapes differ, some workers find plugs uncomfortable and others find muffs hot, and the job may require communication, compatibility with other equipment or protection against impulse noise. Offering a choice of at least two or three types and sizes improves comfort and consistent use, which matter more than the highest rating on the package.

How protectors are rated

Protector ratings are based on laboratory tests in which a panel of listeners wears the protector and the difference between their hearing thresholds with and without it is measured at several frequencies. The results are combined into a single number or class. Because the test conditions are ideal, with careful fitting, the published ratings describe what is possible, not what is typical.

The NRR is still the most widely printed rating in North America. The EPA labeling rule that defines it dates from 1979 and remains in force in 2026; proposals to modernize the test method and label have not been adopted. NRRs are typically 20 to 33 dB for earplugs and 18 to 31 dB for earmuffs.

Many protectors sold in the United States also carry Canadian CSA Z94.2 classes or European SNR values, because the same products are sold in several markets. OSHA's Appendix B to 1910.95 describes how to apply the NRR, and the OSHA Technical Manual and NIOSH guidance describe how to derate it for real-world conditions.

Why real-world protection is lower

Workplace studies that measure protection on workers wearing their own protectors as they normally do consistently find much lower attenuation than the label suggests, especially for earplugs. The gap has several causes: shallow insertion of foam plugs, wrong plug size, earmuff seals broken by glasses or hair, worn cushions, and protectors that move during work. The variation between individuals is also large: some workers achieve close to the label value, while others get almost nothing from the same product.

To allow for this, several derating methods reduce the label rating before estimating protection.

def protected(dba, nrr, method, kind="foam"):
    if method == "label":            # dBA - (NRR - 7)
        return dba - (nrr - 7)
    if method == "half":             # half of (NRR - 7)
        return dba - (nrr - 7) / 2
    if method == "niosh":            # NIOSH type-based derating
        factor = {"muff": 0.75, "foam": 0.50, "other": 0.30}[kind]
        return dba - (nrr * factor - 7)
    raise ValueError(method)

cases = [("foam plug", 32, "foam"), ("premoulded plug", 25, "other"), ("earmuff", 27, "muff")]
dba = 100
print(f"Exposure {dba} dBA")
print("protector         NRR   label   50 %   NIOSH")
for name, nrr, kind in cases:
    print(f"{name:<17} {nrr:>3}  {protected(dba, nrr, 'label'):6.1f} "
          f"{protected(dba, nrr, 'half'):6.1f} {protected(dba, nrr, 'niosh', kind):7.1f}")

Exposure 100 dBA
protector         NRR   label   50 %   NIOSH
foam plug          32    75.0   87.5    91.0
premoulded plug    25    82.0   91.0    99.5
earmuff            27    80.0   90.0    86.8

At 100 dBA, every protector looks excellent on the label but none is clearly adequate after derating. The premoulded plug, with a 70 percent NIOSH derating, offers very little. Results like these explain why engineering controls are essential at high levels and why individual fit testing, which replaces assumptions with measurements, is increasingly used.

Dual protection

At very high levels, an earplug and an earmuff can be worn together. The protection is not the sum of the two ratings, because sound reaches the inner ear through bone conduction, bypassing both protectors. A common rule, used by OSHA, is to add 5 dB to the rating of the more protective device. In practice, dual protection gives roughly 3 to 10 dB more than the better single protector, and bone conduction limits total protection to around 40 to 50 dB at most frequencies. Dual protection is usually recommended above about 100 to 105 dBA and is required by MSHA at very high mining exposures. It should be treated as a signal that engineering controls are urgently needed.

The octave band method

The most accurate laboratory-based method uses octave band measurements of the noise and the protector's octave band attenuation data. For each band, the assumed protection value, typically the mean attenuation minus two standard deviations, is subtracted from the A-weighted band level, and the protected bands are combined. This method shows clearly when a protector is weak at the frequencies where the noise is strongest, which is common for low-frequency noise.

import math

bands = [125, 250, 500, 1000, 2000, 4000, 8000]
a_corr = [-16.1, -8.6, -3.2, 0.0, 1.2, 1.0, -1.1]
noise = [102, 100, 96, 92, 88, 84, 80]          # unweighted band levels, low-frequency noise
mean = [18.0, 20.0, 24.0, 28.0, 32.0, 38.0, 36.0]  # protector mean attenuation (illustrative)
sd = [4.0, 4.0, 3.5, 3.5, 3.0, 3.5, 4.0]

def combine(levels):
    return 10 * math.log10(sum(10 ** (L / 10) for L in levels))

a_bands = [n + a for n, a in zip(noise, a_corr)]
protected = [La - (m - 2 * s) for La, m, s in zip(a_bands, mean, sd)]
print(f"Unprotected: {combine(a_bands):.1f} dBA")
print(f"Protected (mean - 2 SD): {combine(protected):.1f} dBA")
worst = max(range(len(bands)), key=lambda i: protected[i])
print(f"Band contributing most after protection: {bands[worst]} Hz")

Unprotected: 98.1 dBA
Protected (mean - 2 SD): 82.5 dBA
Band contributing most after protection: 250 Hz

For this low-frequency noise, the protected level is controlled by the 125 to 500 Hz bands, where the protector is weakest, while the high-frequency bands contribute almost nothing. A protector with an L rating, or one with better low-frequency performance, would be a better choice.

Over-protection

More attenuation is not always better. A protector that reduces exposure far below what is needed can make speech and warning signals inaudible, isolate the worker and encourage removal. Most guidance suggests aiming for a protected level of roughly 70 to 80 dBA, enough to bring exposure safely under the limit without cutting the worker off from the environment. Workers with existing hearing loss are particularly vulnerable to over-protection, because they may not hear alarms at all. Level-dependent electronic earmuffs, flat-attenuation earplugs, which reduce all frequencies equally, and communication headsets help in these situations.

Selecting protectors for the job

Selection brings together the exposure, the noise spectrum and the needs of the work and the worker. Questions to ask include: How much attenuation is needed after derating? Is the noise low-frequency? Is it continuous, intermittent or impulsive? Does the worker need to hear speech, radios or warning signals? Is it hot, dirty or wet? Does the worker wear a hard hat, safety glasses, a respirator or a welding helmet? Will protectors be put on and taken off frequently? Does the worker have an existing hearing loss or an ear condition? The answers usually point to two or three suitable products, among which the worker can choose.