Assessment of Bias in Measurement of Mercury Emissions from Coal Fired Power Plants – Comparison of Electronic CEMS and Sorbent Traps

Mercury emissions from coal fired power plants are generally measured using one of two methods electronic Continuous Emission Monitoring Systems (Hg CEMS) that most commonly use atomic fluorescence, or sorbent traps. Electronic Hg CEMS are often compared against sorbent traps using EPA’s Method 30B Relative Accuracy Test Audit (RATA). Because sorbent traps include mercury contained in the particulate matter (PM) that escapes the boiler’s PM control device, there is an expected bias of the sorbent trap indicating a higher result than the electronic Hg CEMS. For uncontrolled units this difference is normally an insignificant amount of mercury compared to the total. As will be shown, for controlled units, this difference can be more significant. Some researchers have proposed additional explanations for apparent differences in these methods, to include interference from bromine that may be present in the flue gas. In this effort, work by others is reviewed and comparative data from electronic CEMS and method 30B RATA are presented. The data includes conditions where bromine is injected, where activated carbon is injected and where activated carbon is injected while bromine is introduced to the flue gas. This data is examined for trends in bias and is assessed statistically to determine the confidence levels in any observed trends. The results of the analysis of the data will be used to help test the various proposed explanations for the observed difference between electronic CEMS and sorbent traps and will provide practical guidance for facility owners going forward. 10 IEA Mercury Emission from Coal Workshop Clearwater, FL