Use of Expostats and the UTL 95,70 with ACGIH 3X5 rule

Sampling strats and stats
1

I would like to get opinions on the application of statistics for worker exposure for evaluation of the distributions for shorter term exposures in addition to the 8 hr. TWA evaluations. I’m just started using Expostat and its a great tool to assist in the data analysis and I commend efforts to make it available to the IH community!

For 8 Hr. TWA-OEL calculations and estimations we are going to move to using the EN689 UTL95,70 tests (where we have enough samples) to show compliance or non-compliance.

The issue is most of our exposures are 30 min – 4 hrs. and then zero exposure for the rest of the shift. I think it’s prudent to control any “peak or task duration concentrations” and be simliarly statistically ‘confident’ in the comparison.

As you probably know, the ACGIH 3x5 rule for peak exposure control limits the exposures to between 3x the 8 hr. TWA-TLV for 15-minute periods, 4 times a day separated by an hour and no more than 5x. this is to supplement an 8 hr. TWA measurement or calculation. However, it doesn’t mention the use of statistics or a confidence limit to comply with the 3x5 rules.

My proposal is to use the same EN689 preliminary and UTL95,70 statistical tests and essentially use an exposure multiplier of 3x to the 8 hr. OEL value in EXPOSTAT. This would show we are confident enough to say we are below this 3x ACGIH excursion/peak limit. We would use 5x if the exposure were 15 minutes. this assumes no other STEL or C value exist which is mostly the case for our pharmaceutical dusts.

Has any other company or institution been applying statistics and using the ACGIH 3x5 rule or other exposure multiplier for shorter term peak exposure evaluation? I don’t see any publications and examples where the ACGIH 3X5 rule was used with a statistical comparison. All the text examples are using an 8 hr. TWA and statistics.

Regards

James Castellanos CIH, CSP, MS

2

Some thoughts

It makes perfect sense to me to evaluate the distribution of short term exposures as well as TWA exposures, especially given what you told about when exposures happen.

The ACGIH multiplication factor is a rule of thumb based on an assumed know exposure variability : if a lognormal distribution has a GSD of 2 (maybe 2.5 I don’t remember), few exposures will be above 3GM, and almost none above 5GM. So the reasoning is that, if the GM of the distribution of your short term exposures is just below the TLV-TWA, very few should be above 3*/5* the TLV-TWA if their variability is reasonable.

Because of that definition, linked to variability and not toxicology, I have been reluctant to use these in the absence of a TWA (the multiplication factor in expostats was created when people want to quickly test the influence of changing the oel, not because of the ACGIH rules)

Since your approach seems to be driven by a precautionary principle, I would even not use a multiplication factor, or select an arbitrary threshold value not based on exposure variability considerations.

That said, the ACGIH rules are well known, and it might be easier to convince people to accept them than other values.

I haven’t heard of anyone using 3*/5*OEL with expostats, but not everyone tells me what they do