Basic Exposure Modelling Tools
Exposure modelling is a semi qualitative means of estimating risk associated with particular exposure scenarios. It considers the physical properties of the chemicals of concern, the work being undertaken, controls in place, and environmental conditions influencing exposure levels. The exposure concentration, and the influence factors from exposure scenario conditions are based on historical monitoring datasets.
Exposure monitoring is commonplace in Europe in part due to REACH legislation, as well as in the US. In Australia, it doesn’t have the same ubiquity, which is unfortunate as it it can be useful in:
- Cheaply and quickly assess a large number of exposure scenarios, and apply control banding principles
- Prioritising exposure scenarios for quantitative assessment
- Preemptive assessment of the introduction of chemicals, activities or controls
Ready Tools
There is a plethora of modelling easy, ready to use tools available ranging from extremely simple screening tools (“Tier 1”) to more complex systems with dozens of parameters (“Tier 2”). Some of the most well known models include The Advance Art REACH Tool, IHmod (developed by AIHA), and TREMOX. TREMOX is particularly interesting as it contains multiple popular models within it allowing for comparison, including a bunch of Tier 1 tools.
Pictures are clickable links.
At least one tool (Advanced REACH Tool) allows for Bayesian incorporation of monitoring data into the model to improve reliability. As few as 2-3 samples relevant to the exposure scenario (including historic or research measurements) can drastically improve the precision. Care should be taken however, when using model as an informed prior for your monitoring campaigns. It’s advisable to ensure the statistical prior aligns with your subjective confidence and understanding of the exposure scenario.
The exposure level estimates are unsurprisingly far less accurate than actual monitoring. Various studies estimate a ~2 - 7x difference between modelled and actual exposure levels. Furthermore, models generally overestimate low exposures and underestimate high exposures. These issues may at first seem to render the approach as basically worthless. Evidently modelling is no replacement for monitoring when, say, determining legislative compliance and an exposure standard. The accuracy limitations hasn’t prevented the effective use of real-time technology. It’s dependant on intent and context.
Word of Caution
While these tools seem impressive, be wary of their unfettered use. Many of these ready-use modeling tools are effectively “black boxes”. The foundational exposure monitoring data, and the modeling formulas are not publically available:
Plug in a few catagorical parameters
????
Estimate of exposure
Skeptics suggest that without knowing how these estimates are calculated it is irresponsible to rely on the outputs. Their ease of use has led to a spread of (probably well-intentioned) misinformation about workplace exposures.
Mathematical Models
A far more reliable and transparent method of modelling is to do the calculations yourself (or transparent software). These calculations range from straightforward to intimidatingly complex depending on the assumptions and simplifications you are willing to make.
Mathematical models are fascinating and deserve a full post of their own - one with examples (stay tuned).
What do you think?
I realise this is probably old news to many hygienist readers, but my impression is that exposure modelling appears to be under appreciate tool our hygiene tool belts. Particularly in Australia. I’d be super interested in hearing how you’ve used exposure modelling in your work.