Two Solvents, One Worker, Three Sets of Limits
At a coatings batch-mixing station, personal 8-hour monitoring shows toluene at 62 ppm and methyl ethyl ketone (MEK) at 88 ppm. A 15-minute sample taken during drum changeout shows toluene at 190 ppm. OSHA PELs (no state-plan substitution for OSHA limits) are toluene 200 ppm TWA with a 300 ppm ceiling and a 500 ppm 10-minute peak, and MEK 200 ppm TWA. The state where the plant operates has adopted the ACGIH TLVs as enforceable 8-hour limits: toluene 20 ppm with a skin notation, MEK 150 ppm with a 300 ppm STEL. The company's internal standard is to keep exposures below 50% of the most stringent applicable limit. The plant industrial hygienist reports 'all OSHA PELs met, no action needed.'
1. Compute the mixture exposure index against both sets of limits, evaluate the toluene peak, and assess the hygienist's conclusion.
Read the worked answer
Against OSHA PELs (both solvents are CNS depressants, so OSHA's additive mixture formula applies):
$$EI = \frac{62}{200} + \frac{88}{200} = 0.31 + 0.44 = 0.75$$
Below 1. The toluene peak of 190 ppm is under the 300 ppm ceiling and the 500 ppm 10-minute peak.
Against the state-enforceable ACGIH limits:
$$EI = \frac{62}{20} + \frac{88}{150} = 3.10 + 0.59 = 3.69$$
Nearly four times the mixture limit; toluene alone is 3.1× its 8-hour value. MEK has no measured STEL sample - an additional gap, since the drum changeout is exactly when a MEK spike would occur.
Toluene excursion. ACGIH excursion limits for a TLV-TWA of 20 ppm allow no more than 3× the TLV (60 ppm) for up to 15 minutes, no more than four times per shift, separated by at least 60 minutes. The measured 190 ppm for 15 minutes exceeds that by more than 3×. So the state limits are breached on three counts: the 8-hour toluene value, the mixture index, and the excursion limit.
The hygienist's conclusion: technically correct about OSHA, and wrong as a management decision. The facility is out of compliance with the state-enforceable limits that actually apply, and the OSHA-only view also ignores the skin notation (see next task). The report should be rejected and reissued against the most stringent applicable limits.
2. Translate the internal 50%-of-limit standard into target concentrations and define the minimum control strategy. Address the skin notation and respiratory protection.
Read the worked answer
Targets - and a trap in the wording
Naively, 50% of each limit gives toluene 10 ppm and MEK 75 ppm. But if both sit at their individual 50% targets, the mixture index is 0.5 + 0.5 = 1.0, which breaches the internal standard. The internal standard has to be applied to the mixture index, not per component:
$$\sum_i \frac{C_i}{L_i} \le 0.50$$
One defensible split: toluene 5 ppm (0.25) and MEK 37 ppm (0.25). Starting from an index of 3.69, that is roughly a 7.4× reduction in combined exposure.
Control strategy, in hierarchy order
- Substitution/elimination - waterborne or exempt-solvent coating, or a pre-mixed closed system that eliminates the drum changeout that produced 190 ppm.
- Engineering - closed-loop mixing, drum pumps instead of pouring, local exhaust ventilation at the drum bung during transfer, and enclosure of the batch tank. Most of the peak is a task-based emission (open drum, transfer), which is the easiest thing to capture at the source.
- Administrative - automate the column/level checks, rotate only as a supplement, and never as the sole control where an excursion limit is exceeded.
- PPE - last.
Skin notation. Toluene's skin designation means dermal absorption can add materially to dose, and air sampling does not see it. The true body burden is higher than 62 ppm implies. This also means respirators cannot fully control the exposure. Select gloves on measured breakthrough for toluene (thin nitrile breaks through in minutes; verify against manufacturer data), and change them on a schedule, not when they look worn.
Respirators. A half-mask APR with organic vapor cartridges has an APF of 10, so the maximum use concentration against a 20 ppm limit is 200 ppm. The 190 ppm peak sits inside that - with essentially no margin, and no protection against dermal uptake. Use respiratory protection only as an interim measure under a written program with fit testing and cartridge change schedules, while the engineering controls are installed.
3. A resident 150 m downwind asks about cancer risk. Dispersion modeling predicts an annual average toluene concentration of 0.12 mg/m³ at the fence line. How do you evaluate and communicate this?
Read the worked answer
There is no cancer risk number to compute, and saying so is the honest answer. Toluene is not classified as a human carcinogen; EPA has not derived an inhalation unit risk or slope factor for it. Constructing a cancer risk estimate would require inventing a toxicity value.
What the appropriate assessment is: noncancer hazard.
$$HQ = \frac{C}{RfC} = \frac{0.12\ \text{mg/m}^3}{5\ \text{mg/m}^3} = 0.024$$
An HQ of 0.024 is roughly 2.4% of the level at which no adverse effect is expected - a margin of about 40×. If the site emits other VOCs with the same target organ (CNS, respiratory irritation), sum the HQs into a hazard index; the reference value used should be the most stringent applicable one - some states publish chronic reference exposure levels far below the federal RfC (values in the range of a few tenths of a mg/m³ are common), and using that denominator could move the HQ from 0.02 to something like 0.4. Still below 1, but the difference matters to the answer you give.
Context for the resident
- 0.12 mg/m³ of toluene is about 0.03 ppm. The odor threshold is on the order of 1 ppm, so this is below the level at which most people would smell it.
- The predicted value is a model output, not a measurement. Confirm with fence-line monitoring before making strong claims, and disclose the model's conservatism.
- The meaningful exposure question is not the fence line 150 m away - it is the worker standing at the drum, whose 8-hour TWA is 62 ppm against a 20 ppm limit. Do not let community reassurance distract from the occupational problem that is actually out of compliance.