Assessment of Measurement Uncertainties in a Commercial Atmospheric Mercury Speciation System

Abstract:

The Tekran® 2537/1130/1135 mercury speciation unit is a robust analytical system which is routinely deployed in regional, national, and global monitoring networks to measure concentrations of gaseous elemental mercury (GEM), gaseous oxidized mercury (GOM), and particulate-bound mercury (PBM). Standard operating protocols have been developed to ensure the collection of uniform and high-quality monitoring data. However, recent published laboratory and field research has pointed to the underestimation of GOM using KCl-coated denuders in humid and/or ozone-rich environments. Under very dry conditions, contamination of the Tekran® regenerable particle filter used to collect PBM can lead to the conversion of GEM to PBM, and a subsequent overestimation of reactive mercury (RM) concentrations. In March 2015 we deployed a total mercury (TM)/GEM detector at NOAA’s Mauna Loa Observatory (MLO: Elevation 3397 m) on the island of Hawaii, to calculate reactive mercury (RM = GOM + PBM) as the difference of TM – GEM. We also installed a second Tekran® speciation system to accompany an identical system operating at MLO since 2011. We will present side-by-side comparisons of ambient data collected by the three instruments under a variety of conditions, as well as measured performance parameters assessed using a GOM calibration device. Initial results confirm the degradation of GOM measurement under even moderate levels of humidity, but also suggest that current AMNet sampling and data reduction protocols can exacerbate the problem.