Uncertainty analysis of stack gas flow measurements with an S-type Pitot tube for estimating greenhouse gas emissions using a continuous emission monitoring system

Abstract

In order to reduce greenhouse gas (GHG) emissions from producers in the energy and industrial fields to mitigate climate change more effectively, accurate and reliable studies of methods capable of estimating GHG emissions with proper uncertainty levels should initially be carried out. Continuous emission monitoring systems (CEMS) directly measure GHG emissions by measuring the GHG concentrations and volumetric flow rates of flue gases in smokestacks. Most studies of CEMS have tended to focus on the accuracy and uncertainty levels associated with GHG concentrations rather than on volumetric flow measurements. Particularly, the uncertainty associated with volumetric gas flow measurements by CEMS at a smokestack considering the individual plant must be determined to reduce the overall uncertainty. In the present study, the uncertainties of stack gas flow measurements with an S-type Pitot tube to estimate GHG emissions by CEMS are evaluated by applying the GUM and Monte Carlo methods to an on-site smokestack of an energy power plant. Velocity measurements in the stack were conducted using an S-type Pitot tube mainly used in Korea. The mathematical model equations of the accumulated volumetric gas flowrate measurements for estimating GHG emissions by CEMS are expressed in the form of detailed uncertainty equations by GUM. In the uncertainty evaluations by GUM, the combined standard uncertainty equation is expressed with the average gas flow velocity, stack diameter, static pressure, temperature and water content. The relative expanded uncertainty for volumetric gas flowrate measurements with an S-type Pitot tube in a smokestack by GUM is estimated to be 3.81% for a coverage factor (k = 2) at a confidence level of approximately 95%. In addition, the Monte Carlo method was applied to evaluate the associated uncertainty of volumetric gas flowrate measurements by an S-type Pitot tube for comparison with the results by the GUM method. The results of the uncertainty evaluations by GUM and MCM show good agreement, with only a 0.05% difference in the uncertainty outcomes.