Dylos in Wisconsin
How the Dylos works
The Dylos uses a similar mechanism to the lower cost Shinyei and Syhitech sensors found in the Speck and Dustduino designs under development, only using a laser rather than focused light whose interruption is measured to gauge dust transit. Don Blair did a great job explaining and illustrated the concept of identifying particles by their transit through a beam of light..
How the Dylos is being used
Since 2011 Hank Boschen has been working to setup Dylos monitors, more can be seen under the menu at the top of that page"Laser Air Pollution Monitors." Jeff Falk has also done an analysis of the Dylos monitoring program. PDF link here.
Red markers are active monitoring sites. Click on them to see data:
View Frack Sand Mine Locations, MN and WI in a larger map
Official acceptance of the data has not been forthcoming. Concerns from the Wisconsin DNR (dept. of Natural Resources) and industry groups are that the Dylos monitors use a proprietary calibration, are designed for indoor use, and that the data is just particulates and can't distinguish silica. Distinguishing silica in "fugitive emissions" is not well studied. Drexel University is studying the accuracy of Dylos sensors in comparison to filter-based methods, although I haven't dug into that research yet.
Jeff Falk's Analysis
Jeff starts by noting that wind, humidity, and the differing reflectance of particles are not accounted for by the Dylos, Still, across sites where monitors are up and downwind from a pollution site, the measures will share the same weather effects. The Speck particulate monitor includes a humidity sensor, probably for this very reason.
In his final analysis, Jeff uses only dates where the wind was blowing away from the AOC sand facility, ignores high humidity days (80% or above), and does an analysis with the remaining data. He finds that: "of the 57 days with wind condition and humidity appropriate, 3.5% show possible exceedances of the EPA PM2.5 standard on a 24 hour basis, 37% show possible exceedances of this standard on an average hourly basis, and 51% have at least one hour which possibly exceeds this standard."
He also notes that it is clear that the DNR is not following best monitoring practices and all but admits it in their report. Pretty damning, and backed up by data!