The Exposure Assessment SOP

Purpose

This procedure answers one question for every task the shop performs: how much of what does each person breathe, swallow or absorb doing it, and is that inside the limit. The answer sets the controls, decides who goes into medical surveillance, and becomes a record OSHA expects the shop to keep for decades. Without it, a shop's controls are guesses, and the first evidence of a wrong guess is a blood test or a cough months later. The vocabulary (dose, route, action level, time-weighted average) is owned by a sibling; Dose, Route and the Exposure You Cannot See.

Scope

Covers every recurring task the shop performs, office and field, and every agent those tasks can release: dusts, fumes, vapors, gases, noise and skin-contact chemicals. Covers the initial assessment, monitoring, interpreting results, notifying employees and reassessment.

Does not cover: choosing among controls once a result is in (see related: The Hierarchy of Controls on a Service Call), respirator selection and fit testing (see related: OSHA Respiratory Protection - Annual Fit Test), the medical side of surveillance and its records (see related: The Exposure and Medical Records SOP), or a material discovered mid-job as suspect asbestos, which stops under the shop's asbestos discovery protocol and is not assessed by this procedure.

Roles and responsibilities

The hygienist can be a certified industrial hygienist (CIH) the shop hires, the insurer's loss-control consultant, or the state's OSHA On-Site Consultation Program, which is free for small businesses and kept separate from enforcement.

Procedure

1. Build the task inventory from the job records, not from memory. Pull the last 12 months of job types and templates and list each distinct task (dry-scraping exterior trim, coring a block wall, brazing in a closet, applying a solvent adhesive) with the material it disturbs, the agent it can release, the route (inhalation, ingestion, skin), typical minutes per shift and days per year. Acceptance: every job type run 5 or more times in the year appears, and every row names a specific agent, never "dust" or "fumes". Wrong looks like a row the safety lead cannot attach to an agent; stop rule: it is flagged "material unknown" and the task is not run again until a qualified person identifies the material. Hazard: none at this step, it is desk work; identifying the material later is done by sampling under the asbestos and lead rules, never by breaking it to look.

2. Screen each row against the substance-specific standards. Tag each row with the standard that applies, naming the Part: lead, 29 CFR 1926.62 (construction) or 1910.1025 (general industry); respirable crystalline silica, 1926.1153 or 1910.1053; noise, 1910.95 for general industry, with the construction duty at 1926.52; plus any agent in the air contaminant tables at 1910.1000. Acceptance: every row reads "standard applies", "screened out" with a written reason, or "unknown". Wrong looks like "screened out" with no reason, or a row tagged to the general-industry Part for work done in a customer's house under construction; stop rule: an unknown row goes to the hygienist within 10 working days (a shop default), and until then the task runs with the shop's most protective control for that material type. Hazard: none, desk work.

3. Use a presumption, a compliance table or objective data before paying for sampling. Three things can stand in for monitoring. Fully following the specified controls for a task in Table 1 of 1926.1153 relieves the construction employer of the silica exposure assessment for that task. Under 1926.62(d)(2), named lead tasks (among them manual scraping, manual sanding, heat gun use and power tool cleaning with dust collection, on lead paint) are presumed above the PEL, and the employer provides interim protection, respirators included, until monitoring shows otherwise. Objective data, records showing a task cannot release an agent above the action level under foreseeable conditions, can also serve where the standard allows it. Acceptance: each row records its basis: Table 1 row named, presumption cited, objective data source with date and conditions, or "needs monitoring". Wrong looks like objective data borrowed from different conditions (outdoor numbers for an indoor task, a vacuum-shrouded tool for a bare one); stop rule: conditions that differ make the row "needs monitoring". Hazard: none, desk work; the presumed-high rows go out under interim protection from today, not from the day results arrive.

4. Run personal sampling on the "needs monitoring" rows. The hygienist puts a sampler on the worker doing the task, with the inlet in the breathing zone, for the full shift or the full task, on a day that represents the task's worst realistic conditions; samples go to a lab accredited for that analysis. Noise uses a dosimeter worn the whole shift. Acceptance: the sampling record states the worker, task, minutes on task, controls actually in use, indoor or outdoor, and room size. Wrong looks like a sample from a different task or a day when the planned controls were not in use; stop rule: the sample is kept and labeled with what was actually done, never discarded, and the planned sample is rescheduled within 30 days. Hazard: the sampled worker keeps every control the shop currently requires, interim respirators included; a monitoring day is never the day to take the respirator off "to see the real number", and the sampler tubing is routed under clothing so it cannot catch on a spinning tool.

5. Compare each result to both limits and set the disposition. Write each result as an 8-hour time-weighted average beside the OSHA action level, the OSHA permissible exposure limit (PEL) and the NIOSH recommended exposure limit (REL), which is often stricter. Lead under 1926.62: action level 30 and PEL 50 micrograms per cubic meter of air. Silica under 1926.1153: action level 25, PEL 50. Noise: 85 dBA triggers hearing conservation in general industry under 1910.95(c), while the construction rule, 1926.52, keys its program to exposure above its 90 dBA table; NIOSH recommends 85 dBA with a 3 dB exchange rate. Where the REL is stricter, the shop uses it as its own trigger. Acceptance: every result carries one disposition: below action level, document and reassess on change; at or above action level, periodic monitoring and medical surveillance as the standard requires (for lead in construction, 1926.62(d)(6) sets every 6 months between action level and PEL and every 3 months above the PEL); above the PEL, engineering and work-practice controls plus respirators under the written program. Stop rule: a result above the PEL pauses that task the same day until controls change, rather than waiting for the next monitoring cycle. Hazard: none at the desk; the hazard is the crew still doing the task on the old controls, which is why the pause is same-day.

6. Notify each employee in writing. Give each sampled worker, and each worker the result represents, a written notice of the result, the limit, and the corrective action when it exceeded the PEL. The construction lead and silica standards set five working days after the shop receives results; check the text of the standard that applies, since the general-industry silica standard gives 15. Acceptance: a dated copy in the assessment file, signed or acknowledged by each worker. Wrong looks like a result posted on a board in a shop where half the crew never comes in; stop rule: any worker not reached in the window gets the notice by hand or mail the next working day. Hazard: none, desk work.

7. Reassess on any change, and at least yearly. Triggers: a new task, material, tool or product; a task moving indoors; a big change in minutes per shift; a symptom report; a medical result above the shop's trigger. Acceptance: the inventory carries a reviewed date no more than 12 months old, and each trigger is closed within 30 days. Wrong looks like a cough report that changed nobody's controls; stop rule: a symptom report pauses that worker's exposure to the suspected task until the safety lead and the physician have looked at it. Hazard: none at the desk; the worker who reported symptoms is the one at risk.

The record this produces

One assessment file per inventory row: the task, agent, route, basis (presumption, Table 1 row, objective data or sample), sampling record, lab report, the computed TWA beside each limit, the disposition, the employee notice, and the next review date. Under 29 CFR 1910.1020 exposure records are kept 30 years and medical records for the duration of employment plus 30 years, so they live in the shop's permanent files, not a tech's phone. They are read later by the safety lead setting controls, by a physician evaluating a worker, by an OSHA inspector, and by a worker or their representative, who have a right of access under 1910.1020.

Worked pass: a nine-person repaint and repair shop

Step 1. The safety lead pulls the year's jobs: 14 task rows. Three matter here: interior window-sill sanding on pre-1978 houses (about 38 jobs), cutting fiber-cement siding (22 jobs), and a demolition hammer on bathroom tile (11 jobs). Every job type run 5 or more times appears.

Step 2. Sill sanding: lead, 1926.62. Fiber cement: silica, 1926.1153. Tile demolition: silica and noise.

Step 3. Fiber cement is cut outdoors with a saw and dust collection that matches its Table 1 row, so that row is closed without sampling. Sill sanding is shrouded power sanding on a HEPA vacuum, which 1926.62(d)(2) presumes above the PEL as power tool cleaning with dust collection, so the tech already works under interim protection, respirator included; it is "needs monitoring" because only a result can lift that. Tile demolition is "needs monitoring" for noise.

Step 4. This step fails. On the planned sill-sanding day the shroud's vacuum hose split at 10 a.m. and the tech finished by hand-sanding dry, also a presumed-high task, wearing the interim-protection respirator. The sample, 4 hours on task, came back at 84 micrograms per cubic meter. The stop rule applies: the sample is kept and labeled "dry hand sanding, shroud failed", not "sill sanding", and the planned sample is rescheduled. (Stopping the sanding when the hose split belonged to the dusty-task procedure; see related: The Dusty Task Setup SOP.)

Step 5. As an 8-hour TWA with no other lead exposure that shift: 84 times 4 hours divided by 8 hours is 42 micrograms per cubic meter. That sits between the action level of 30 and the PEL of 50. Because the worker was exposed at or above the action level on a day, 1926.62(j)(1)(i) calls for initial blood lead testing for him. The rescheduled shrouded sample, 4 hours on task, came back at 18, a TWA of 18 times 4 divided by 8, or 9: below the action level of 30, so the presumption's interim protection can be lifted for that task as performed, documented, reassess on change. The tile-demolition dosimeter read 88 dBA over the shift: under the 90 dBA construction table, but over the 85 dBA NIOSH REL the shop uses as its trigger, so the two tile workers enter its hearing conservation program.

Step 6. Results reached the shop on a Monday; written notices went to the sanding tech and the two tile workers by Thursday, inside five working days.

Step 7. The review date is set 12 months out, and a new trigger is added to the inventory: any shroud or vacuum failure on a lead task opens a reassessment.

Reading the general section against this pass: the presumption in step 3 covered the dry hand sanding the moment it happened, so the respirator the tech was wearing when the hose split was the required control, not a lucky one, and the 42 is a real exposure that went into his record rather than a bad sample to throw away.

References

• OSHA 29 CFR 1926.62 and 1910.1025, Lead (exposure assessment, interim protection, monitoring frequency, notification)
• OSHA 29 CFR 1926.1153 and 1910.1053, Respirable Crystalline Silica (Table 1; exposure assessment)
• OSHA 29 CFR 1910.95, Occupational Noise Exposure; 29 CFR 1910.1020, Access to Employee Exposure and Medical Records
• NIOSH Pocket Guide to Chemical Hazards and Criteria for a Recommended Standard: Occupational Noise Exposure
• See related: Dose, Route and the Exposure You Cannot See