Two Compliance Clocks on One Rinse Line

A result above the PEL is not a statistical curiosity to be argued away, but one sample is also not a characterization of a six-person group. Both things are true and the action plan has to reflect both.

What I can conclude:

• 6.9 versus 5 is a 38% exceedance. Typical sampling and analytical error for Cr(VI) methods is on the order of 25%, so a false positive driven by analytical error alone is unlikely. The result is probably real for that operator on that day.
• The engineering evidence corroborates it. The local exhaust is at 55 fpm against a 100 fpm design and the duct is corroded through near the scrubber. A capture velocity more than 40% below design at the tank rim is a known cause of chromic acid mist escaping into the breathing zone. Two independent lines of evidence pointing the same way outweigh the small sample size.

What I cannot conclude: the exposures of the other five operators, the night shift, or peak/ceiling conditions during tank loading or line maintenance. The voided cassette means half my intended sample set does not exist.

What I do now, before resampling:

• Treat all six as potentially overexposed. Fit-test and issue respirators as an interim measure and start the medical surveillance and exposure-notification machinery.
• Repair or replace the exhaust. This is a known deficiency with a known fix; it does not require more data.
• Resample properly - multiple personal samples covering all operators, both shifts, and the high-exposure tasks, with cassette handling checked before the samples go out. The voided seal is a sampling-protocol failure and should be corrected at the source, not just noted.
• Do not assume the sample is unrepresentative because the operator was 'doing something unusual.' Ask, document, and then decide.

The hazard is a carcinogen with no established safe threshold, delivered by inhalation of mist and by dermal contact with solution and contaminated surfaces. That shapes the ranking.

1. Elimination or substitution. If the customer's specification allows trivalent chromium or a non-chromium coating, this is the only control that removes the hazard. Worth investigating even if the answer is no, because 'we looked and the spec forbids it' is a defensible record.
2. Engineering controls. Restore and upgrade the local exhaust: repair the corroded duct, rebalance the system, verify capture velocity at the tank rim against design, and enclose or baffle the tank. Consider a mist suppressant - but check its chemistry first, since some fluorosurfactant suppressants are PFAS and can create a separate regulatory and environmental problem.
3. Work practices and housekeeping. Wet methods and vacuuming rather than dry sweeping; the standard requires housekeeping to prevent accumulation of Cr(VI) dust on surfaces. Written procedures for tank loading and for maintenance inside the line.
4. PPE. Respirators and chemical protective clothing as interim protection until repeat sampling demonstrates the engineering controls are working. Also change rooms, hygiene facilities, and washing before breaks and at the end of shift, because ingestion and take-home exposure are real pathways.

What I would not rely on:

• Dilution or general ventilation. It does not capture mist at the source.
• Rotating operators to 'share' the exposure. An 8-hour TWA does not fall because different people take the dose, and Cr(VI) is a carcinogen.
• Respirators as the permanent answer. They are the right interim measure and the wrong long-term strategy, and they depend on fit testing and program discipline that a 40-person shop may not sustain.
• The single sample as proof that everyone is overexposed, or the voided sample as proof that nobody is. Both overreach.

The discharge has happened and cannot be undone. What can still be controlled is everything after it, and the biggest risk is that someone tries to make the record look better than the event.

Immediate:

• Notify the control authority. Industrial user permits almost universally require prompt oral notification of a violation, often within 24 hours, followed by written notice within a short window (commonly five days). The batch went out at 06:10 and the clock is running. Notify today, describe what happened, what was discharged, and what is being done.
• Do not dilute, blend, or re-sample to get a favorable number. That converts a permit exceedance into a falsification problem and is a far more serious matter under the Clean Water Act.
• Preserve the batch composite result and the operating log. Chart the treatment step and the valve timeline; get the night shift's account in writing.

Investigate:

• Was the valve opened because the batch tank was full and the next cycle was starting? That is a capacity and scheduling problem dressed up as an operator error.
• Is the treatment step bypassable at all? A valve that can send untreated rinse water to the sewer without a physical or interlocked barrier is a design defect. Lockable valves, level alarms, and an interlock that prevents discharge before treatment completion are the corrective actions.
• Was the operator trained on the permit limit and on the prohibition on discharging to make room? If not, this is a management failure.

The link to the air side: the corroded duct and the prematurely opened valve are the same story - deferred maintenance and production pressure. The corrective action that covers both is a list of permit-critical and safety-critical equipment with defined inspection frequency and a rule that production scheduling cannot defer work on that list. Reporting the exceedance honestly and early is also the best protection the shop has, given a clean prior record.