Similar exposure groups: how they set the price of air monitoring

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.

Every air monitoring quote you receive is really a price on a number of samples. That number is set by how many similar exposure groups (SEGs) the work breaks into, not by how many people are on the payroll.

A similar exposure group is a set of workers who do the same job the same way. Two plants with 60 workers each can differ four times over in sampling cost, because one breaks into 3 SEGs and the other into 12.

What actually defines an exposure group

An SEG is a set of workers whose exposure to one specific substance is expected to be alike enough that a sample on any one member can stand in for the rest.

Six factors determine what "alike" means:

• Task. What the person does with their hands.
• Material. The specific alloy, coating, resin or ore, not the broad category.
• Process. MIG welding or stick welding, wet cutting or dry cutting, closed transfer or open transfer.
• Location. How close the worker is to the source and to the ventilation that serves it.
• Shift. Day and night crews rarely run the same product mix or the same staffing.
• Time on the task. The share of the shift the worker spends on the task that produces the exposure.

Job title, department and the org chart do not determine it. Payroll titles exist for pay and supervision, and they line up with exposure only by accident.

Here is what that looks like in a 27-person fabrication shop that is sampling for welding fume:

That makes six exposure groups where the org chart showed three, and two of those six need no sampling for this substance. If you group by title, you buy samples on people who are not exposed and learn too little about the ones who are.

Why one sample tells you nothing

Occupational exposure is not a fixed property of a job. It varies randomly from day to day, and the spread is wide.

The same worker, doing what looks like the same job two days in a row, can produce results that differ by a factor of two or more. The causes are ordinary:

• The production rate
• The material
• The weather
• Whether a door was open
• Whether the make-up air unit was running
• How the respirator or the fan happened to be positioned

A single sample gives you one draw from that range. It cannot tell you the average, and it certainly cannot tell you the high end, which is what decides whether you are over a limit on a bad day.

There is no magic number. The count depends on how many distinct jobs there are and how much the work varies from day to day. Highly variable operations need more samples, while tightly controlled, continuous processes can often be covered with a few, and the number is settled after a walk of the floor.

Worst-case or random sampling: choose one deliberately

Both are sound designs. The mistake is doing one and reporting it as the other.

Worst-case (targeted) sampling deliberately selects the most exposed workers, tasks and days. It is an efficient design for compliance screening, because if the worst case is clean, the group is clean. It cannot be used to estimate a group average, because the samples were not selected at random.

Random sampling selects workers and days without bias. It supports statistics: the average, the spread (the geometric standard deviation) and an upper bound (the upper tolerance limit). It is what you want if the goal is to characterize what workers are actually exposed to, or if the results will feed a health surveillance program or an epidemiology study.

A report that presents targeted samples and then calculates a group average from them makes a claim its own data cannot support. So whichever you choose, the design choice and the reasons for it belong in the report.

→ How to read an industrial hygiene report

When the standard's minimum is enough

This is worth stating plainly, because it will save some readers money.

Several of OSHA's substance-specific standards set a compliance minimum that is far below what full statistics on an exposure group would need. The lead standard, for example, requires at least one full-shift personal sample for each shift for each job classification in each work area. That is a floor for demonstrating compliance, not a design for understanding exposure.

If your only goal is to close out an item on an OSHA citation or to satisfy an initial monitoring requirement, the standard's minimum is the standard's minimum.

You want more than the minimum if your goal is one of these:

• To know where you actually stand
• To support a decision not to enroll people in medical surveillance
• To build a record that will survive a dispute years from now

They are different products bought for different reasons, and a written scope of work should state plainly which one it is designed to deliver. So ask which one you are buying.

→ Area vs personal air sampling · → Objective data and exposure assessment · → Air monitoring surveys