Inhalation Exposure from Spills Modeling – Part II

A few years ago I was fortunate enough to be hired by Dr. Perry Logan of the 3M Company to do some solvent spill modeling.   Perry is a great collaborator and this partnership resulted in a considerable amount of experimental work and a paper published online and hard copy as an “open access” document in the Annals of Occupational Hygiene.   For me the really neat part about this work was that the laboratory experiments were done by very savvy technologists and the data generated really showed the validity of the modeling approach.  I would be happy to send a pdf of this publication and the Word document of the extensive supplementary material published online (mostly laboratory test details) to anyone who asks at: mjayjock@gmail.com

During this work we looked at evaporating liquids with constant sources such as might occur from an open vessel or deep pool. We also examined first-order decreasing sources such as typically thin spills as they initially spread out on the floor or counter top.  The first type of source (i.e., constant) would be generally applicable to "non-spill" scenarios with relatively deep evaporating liquids.

Some of the important learnings from this work for me are:

These solvents evaporated so rapidly after being spilled that the exposure was essentially a bolus occurring in a very short time frame (minutes).   As I have written in previous blogs, bolus exposures are not often or well considered by the IH and risk assessment community within most workplace scenarios.

Given no STEL or Ceiling value for any particular compound of interest, one was compelled to use the ACGIH Excursion limit as outlined in the TLV Booklet:

“Excursions in worker exposure levels may exceed 3 times the TLV-TWA for not more that a total of 30 minutes during a workday, and under no circumstances should they exceed 5 times the TLV-TWA, provided that the [8 hour] TLV-TWA is not exceeded.”

In the case of these very volatile solvents this de facto exposure limit was exceeded during 1 L spills by all of the solvents under consideration because they evaporated so quickly. 

Given that situation, evacuation is the most appropriate response to a spill with a quantity beyond a determined threshold.   That threshold can be estimated using the techniques described in this paper and supplementary material.

The practical conclusions from this work of potential use to practicing Industrial Hygienists are presented below:

Clearly, the chamber methods used in this work will not be available to all Industrial Hygienists.  However, the critical factor for the determination of both constant and first-order decreasing evaporation rates for single-component liquids resides in the simple gravimetric measurement of evaporative solvent weight loss with time. The practicing hygienist should consider conducting gravimetric measurement of evaporation rates for both types of models (constant rate for constant sources and decreasing rates for spills) using tools that should be readily available to most investigators. This would involve placing a small container on a scale accurate to within 0.1 g in a laboratory hood. The liquid of interest would be placed in the container and mass loss recorded versus time. The type of scenario to be modeled would determine the specific experimental setup. Constant emission sources, such as an open vessel with evaporating solvent could be simulated with a petri dish or bottle cap with a small surface area to volume ratio in an attempt to produce a constant rate of evaporation for a constrained surface area. Simulation of a spill (an exponentially decreasing source) might use a petri dish, paint can lid, or floor tile in which the solvent thickness represents a realistic value by using a spill volume, which barely reaches the perimeter of the selected surface area. For the constant evaporation rate experiments, given a known surface area and representative air movement over the liquid (measured with an anemometer), one could readily calculate the evaporation rate per unit area. For the exponentially decreasing sources, alpha (α) could also be calculated. The details of both calculation techniques are described in the available supplementary material.

This blog lives on your questions and your experience as exposure assessors relative to the application of modeling techniques.  It will ultimately slow and stop without this participation.  Please let me know what you think about the above information, whether you might consider doing this and what would keep you from it.   That could, in turn, help me to determine what information or specific approach or calculation advice, if any, might be needed.