Solvents and VOCs: sampling for toluene, xylene and MEK

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.

This page covers solvents and other volatile organic compounds (VOCs), including methyl ethyl ketone, known as MEK.

A common mistake in solvent sampling is collecting MEK on a coconut shell charcoal tube.

Charcoal is the standard sampling medium for solvent vapors, and for good reason. It works for the aromatic solvents benzene, toluene and xylene, and for most of the aromatics and chlorinated solvents you will encounter. MEK is the exception.

OSHA method 5004 calls for two 600 mg silica gel tubes in series, and OSHA method 1004 uses different sampling media again:

• Anasorb CMS, a carbon molecular sieve, 150 mg / 75 mg
• The SKC 575-002 passive sampler with Anasorb 747
• The 3M 3520 organic vapor monitor with charcoal pads

The ordinary coconut shell charcoal tube used for the aromatics is not on either list. If MEK is sampled on it, the lab recovers too little of it, by an amount no one can predict, and the reported result is lower than the real exposure.

Suppose you run a shop where MEK and toluene are both in use, as in printing, adhesives, coatings and composites. That means two different sampling trains on the same worker, not one tube and not one pump.

Where we find them

• Spray finishing and coating: booths, prep decks, touch-up, and aerospace and automotive refinishing
• Adhesive and laminating work: contact cement, composite layup and converting
• Screen and flexographic printing: press wash, ink thinning and blanket cleaning
• Parts washing and vapor degreasing: the tank, and the drain-and-refill
• Composite and fiberglass work: resin, gelcoat, styrene and cleanup solvent
• Paint and ink manufacturing: mixing, thinning and pouring from drums
• Maintenance and tool cleaning benches. These are often left out of the sampling plan, and they often produce short bursts of high exposure.

Exposure limits

The table shows limits from three sources: OSHA's permissible exposure limits (PELs), NIOSH's recommended exposure limits (RELs), and the threshold limit values (TLVs) published by ACGIH. The PELs are legal limits, while the RELs and TLVs are recommendations.

All values are in parts per million (ppm). In the table, “TWA” means a time-weighted average over a full work shift, and “STEL” means a short-term exposure limit.

Four points in that table deserve a closer look.

Toluene is in a different OSHA table, Table Z-2, and that changes how the sampling has to be planned. Table Z-2 gives toluene three limits:

• An 8-hour TWA
• A ceiling
• A maximum peak that is allowed above the ceiling for a limited time

A full-shift TWA cannot show that you meet a ceiling, so if toluene is in your process at levels that matter, short-term sampling is the only way to evaluate two of the three limits.

Xylene and MEK have an OSHA 8-hour TWA and nothing else. There is no federal STEL at all. That does not mean short bursts of exposure are unimportant, only that the federal limit OSHA can enforce says nothing about them. The STELs quoted for both come from NIOSH and ACGIH, not OSHA, and any report that uses them should say which is which.

The ACGIH TLVs are lower than OSHA's PELs across the table. Toluene and xylene are both at 20 ppm, against OSHA's 200 and 100, which is 5 to 10 times lower. MEK's gap is smaller but real, at 75 ppm against 200, or about 2.7 times lower. ACGIH also has a 150 ppm STEL for MEK where OSHA has none.

These comparisons are close to like-for-like, because both sides are 8-hour TWAs of the same substance. That is unlike manganese, where the limits cover different particle size fractions. The health effects behind ACGIH's changes to these limits are harm to the nervous system and, for toluene, harm to a developing child.

MEK also carries a “Skin” notation from ACGIH. That notation concerns a route of exposure that air sampling does not measure, which is covered below.

→ OSHA PELs compared with ACGIH TLVs

How we sample for them

Mixed-solvent atmospheres are the real problem, and they are the normal case. A charcoal tube has a limited number of sites where vapor can be held, and the solvents in your process compete for them. A solvent at a high concentration will displace one at a lower concentration that is already held there.

The solvent you care about can therefore be pushed off the front section and out of the tube, while a solvent you do not care about takes its place.

The result is that in a heavily loaded booth or press room, the sample volume has to come down, not up. That is counterintuitive, because you would expect more air to let the lab measure smaller amounts. Here, more air can mean breakthrough, where vapor passes all the way through the tube.

A result with breakthrough tells you only that the exposure was at least that high, and it is not a true measurement. This is why the lab analyzes the back section of a tube separately, and why the report should show both the front and back section results.

Note the air volume for OSHA method 5000 as well. 12 liters at 0.05 L/min is four hours. A full-shift TWA on that method means consecutive tubes, with the results weighted by air volume, not one tube left on for eight hours.

→ How many air samples do you really need?

Skin absorption, and the sample that does not measure it

Air sampling measures what a worker inhales, but with several common solvents, absorption through the skin adds a real share of the dose the body takes in. For a worker who wipes parts by hand with solvent-soaked rags, a passing air result is not the reassurance it appears to be.

MEK is the example here. The ACGIH “Skin” notation on MEK says that skin contact can add a meaningful amount to the total dose. MEK is also exactly the solvent that ends up on rags and bare hands in a print shop or a layup bay.

The report should say so where the work practices call for it. An air monitoring survey cannot measure skin absorption, but the notation belongs in the record, because the person reading the report a year from now needs to know where the survey's reach ends.

Benzene is the exception on this page

Toluene, xylene and MEK are covered only by OSHA's general tables of air contaminant limits. Benzene is not. It has its own OSHA standard, and that standard carries requirements the other three do not have.

The action level shown in the table is set below the PEL, and at or above it the benzene standard requires periodic monitoring.

Benzene enters solvent work as a contaminant more often than as an ingredient, arriving in petroleum naphtha, mineral spirits, some reclaimed solvent and some feedstock made from crude oil. If it is present, it shapes the whole sampling plan, because a 0.5 ppm action level is a different sampling problem from a 200 ppm TWA.

→ Benzene air monitoring

What the rest of a solvent survey does not require

Toluene, xylene and MEK have no substance-specific OSHA standard, so their limits come with no action level, no required monitoring schedule, no deadline for notifying workers and no required medical surveillance.

The schedule is therefore yours to set and to defend, and in practice the trigger is change:

• A new coating system
• A formulation change
• A change to a booth
• A change in shift length
• A complaint

If your process and materials are stable and a well-planned survey showed comfortable margins, annual re-monitoring may be more than you need, and we will say so.

→ What to do when results are above the PEL