Noise Dosimetry and Hearing Conservation in Stone Fabrication

Ask any experienced fabricator whether the shop is loud and the answer is yes, followed immediately by a shrug. Ask what the eight-hour time-weighted average is at the polishing line and the room goes quiet. That gap is the entire subject of this article. Hearing loss from occupational noise is gradual, painless and permanent, it does not announce itself the way a cut or a strain does, and the people most exposed are usually the ones who have stopped noticing. Regulators do not accept impressions, and neither should a shop owner who is signing the payroll.

Measuring noise properly in a stone shop is not complicated, but it is specific. You need to know the difference between what a sound level meter tells you about a machine and what a dosimeter tells you about a person, you need the instrument settings that match the federal rule rather than the factory defaults, and you need to understand that a hearing protector's printed rating is a laboratory number that will not survive contact with a wet, dusty production floor. This guide walks through measurement, the regulatory framework, the controls that work in a fabrication environment, and how to build a written programme that holds up.

Dose Is Level Multiplied by Time

The central concept is dose, not loudness. A worker's exposure depends on how loud each activity is and how long it lasts, summed across the shift. Under the federal occupational noise standard the permissible exposure limit is 90 dBA as an eight-hour time-weighted average, and the standard uses a 5 dB exchange rate: every 5 dB increase halves the allowable duration. Ninety dBA is permitted for eight hours, 95 dBA for four, 100 dBA for two, 105 dBA for one, 110 dBA for thirty minutes and 115 dBA for fifteen. Impulse or impact noise must not exceed 140 dB peak sound pressure level.

A separate and lower threshold governs the hearing conservation programme itself. At an eight-hour time-weighted average of 85 dBA, known as the action level, the employer has to run a full programme: monitoring, audiometric testing, hearing protectors, training and records. The 85 dBA action level is the number that matters most in fabrication, because a shop can sit comfortably under the 90 dBA limit and still owe every one of those obligations. Two thresholds, two sets of duties, and only one of them is a ceiling.

The federal occupational health research institute recommends a more protective criterion: 85 dBA as an eight-hour average using a 3 dB exchange rate, on the reasoning that acoustic energy doubles every 3 dB. Its risk assessment puts the excess risk of material hearing impairment after a forty-year working life at roughly 8 percent at 85 dBA, against roughly 25 percent at the enforced 90 dBA limit. A shop targeting the recommended criterion rather than the legal minimum is buying a substantially different outcome for its long-service staff.

Area Sampling Versus Personal Dosimetry

What a sound level meter can and cannot tell you

A sound level meter measures the sound pressure at a point in space at a moment in time. Point it at a bridge saw during a cut and you learn what that machine contributes at that distance. This is how you rank machines, verify that an enclosure is working, find the reflective wall that is doubling back onto a workstation, and decide where to spend money on controls. Area sampling is diagnostic work.

What an area survey cannot do is establish anyone's exposure, because people move. A stone shop is not a production line where an operator stands in one spot for eight hours. The same fabricator cuts, carries, grinds, polishes, walks to the crib, waits for a slab and spends an hour templating off site. Reconstructing a dose from spot readings and a guessed time budget produces a number nobody can defend in an inspection or a claim.

What a dosimeter tells you

A personal noise dosimeter is worn for the shift with the microphone on the shoulder, in the hearing zone, and it integrates everything the worker actually experiences. It returns the time-weighted average, the accumulated dose as a percentage, and on most modern units a time history that shows exactly which parts of the day drove the result. That time history is often more valuable than the average, because it converts an abstract exposure into a specific task the shop can change.

Sample real days, not convenient ones. Pick workers whose tasks represent the range in the building, run several shifts across different job types, and include the fabricator who does the awkward mix of cutting and hand finishing rather than only the operator who stands at one machine. If nobody is close to the action level on a busy day, the finding is meaningful. If sampling only ever happens on a slow Tuesday, it is not.

Instrument settings that must match the rule

Default settings will quietly invalidate a survey. For compliance measurement, dosimeters should meet the relevant American National Standards Institute specification for personal noise dosimeters at Type 2 accuracy or better, and sound level meters the corresponding standard for sound level meters, also Type 2 or better. Configure A-weighting, slow response, a 5 dB exchange rate, a 90 dBA criterion level and an 80 dBA threshold, so that everything at or above 80 dBA is counted toward the exposure.

Mapping the shop operation by operation

Published noise figures for specific machines vary enormously with blade, water flow, enclosure, room acoustics and how hard the operator pushes, so treat any single quoted number for a saw or grinder as a starting hypothesis rather than a specification. What does transfer between shops is the character of the noise each operation produces, because character determines which control is worth trying. The table below is a planning aid for a walkthrough survey, not a substitute for measuring your own building.

The Compliance Framework in Practice

Audiometric testing

Once a worker is exposed at or above the action level, the employer must establish a valid baseline audiogram within six months of that first exposure, extended to one year where a mobile testing van is used. Baselines should follow a quiet period so that a temporary shift from the previous shift is not baked into the reference. Every covered worker then receives an annual audiogram compared against that baseline.

The comparison looks for a standard threshold shift, defined as an average change of 10 dB or more at 2000, 3000 and 4000 hertz in either ear relative to the baseline. A shift triggers action: notify the employee in writing, refit or upgrade hearing protection, and evaluate whether the exposure that caused it can be reduced. Separately, if a work-related standard threshold shift is accompanied by a total hearing level of 25 dB or more above audiometric zero averaged across those same frequencies in the same ear, the case is recordable on the injury and illness log.

Training and records

Training is annual for everyone in the programme and has to cover the effects of noise on hearing, the advantages and disadvantages of the protector types offered, how to fit and care for them, and the purpose and procedures of audiometric testing.

Recordkeeping is specific and easy to get wrong. Noise exposure measurement records must be retained for two years. Audiometric test records must be retained for the duration of the affected worker's employment, which for a long-service fabricator means decades of files. All of these records must be made available to employees, former employees and their designated representatives on request.

Engineering and Administrative Controls

Where exposures exceed the permissible limit, feasible engineering or administrative controls are required, and personal protection is what remains after those controls have been applied rather than a substitute for them. In fabrication the highest-return control is usually separation: enclosing or curtaining the saw bay, moving the operator station away from the cutting envelope, and keeping hand finishing out of the acoustic shadow of the heavy machines.

Machine-level controls come next. Damped or laminated blade cores, wheels matched to the material, well-maintained spindle bearings and consistent water delivery all change what the machine radiates. Water is doing acoustic work as well as thermal and dust work: a dry start on a core bit or a starved cut produces both a squeal and unnecessary tool wear. Compressed air is a frequent and easily fixed offender, since engineered nozzles and lower supply pressure remove a great deal of high-frequency jet noise.

Administrative controls are the cheapest and the most often forgotten. Rotating hand-finishing tasks between operators spreads dose instead of concentrating it. Scheduling the loudest cutting when fewest people are nearby lowers everyone else's exposure. Building the programme around quiet break areas gives the ear genuine recovery time. None of these reduce the sound the machine makes, and all of them reduce the dose the person receives.

Hearing Protector Selection and Honest Derating

The rating printed on a package comes from a laboratory fit on trained subjects and overstates what a fabricator gets in a wet, dusty shop with gloves on. Under the standard's own guidance, when exposure is measured in A-weighted decibels you subtract 7 dB from the rating before subtracting it from the measured average. Compliance practice frequently applies a further 50 percent safety factor beyond that adjustment, which is where the widely used rule of taking the rating, subtracting 7 and halving the remainder comes from.

The research institute's derating scheme is more granular, cutting the rating by 25 percent for earmuffs, 50 percent for formable foam plugs and 70 percent for other plug types. Whichever scheme you use, the arithmetic must show that protection brings the wearer to at least a 90 dBA time-weighted average, and to 85 dBA or better for any worker who has already shown a standard threshold shift.

Selection is not purely a numbers exercise. Workers must be offered a choice from a variety of suitable protectors, and the employer has to provide initial fitting and supervise correct use. Over-protection is a real failure mode: a fabricator who cannot hear a shouted warning or the change in a saw's note will pull a plug out, and an earmuff worn over safety glasses with dusty hands loses far more attenuation than any derating factor accounts for.

Building a Programme That Survives Contact With the Shop

The written programme should name who owns it, list the sampling schedule and the last results, define the protector options and their derated ratings, describe the audiometric provider and the referral path for shifts, and record training dates. Keep it short enough that a supervisor will actually read it. A binder that documents a programme nobody follows is worse than no binder, because it establishes what the shop knew it should be doing.

Resurvey whenever the shop changes, and stone shops change constantly. New equipment, a relocated polishing line, a change in material mix toward harder engineered stone, a rise in production hours or an added shift can all move exposures. Engineered stone also requires diamond tooling rated for engineered stone, and switching tooling to suit a material change is exactly the kind of alteration that justifies a fresh round of dosimetry.

Run the noise programme alongside the silica programme rather than as a separate exercise. The two hazards overlap almost perfectly in this industry: the same handheld grinding and dry finishing tasks drive both, and the respirable crystalline silica limit of 50 micrograms per cubic meter as an eight-hour average, with an action level of 25 micrograms per cubic meter, applies to precisely the operations that dominate the dosimetry results. Wet methods and enclosure serve both objectives at once.

Tooling choice sits upstream of every measurement in this article. Review current blades, wheels and pads at Dynamic Stone Tools, and use the reference material at dynamicstonetools.com when you are matching tooling to a material, since a correctly specified blade run wet is generally quieter and longer-lived than an overloaded one.