Remote Gas Emission Flux Sensing in a Thermal Power Plant Using Shadow Image Processing

Abstract

Total flue gas emission is a crucial quantity for control of environmental impact in thermal power plants. Direct gas flow measurements by Pitot tubes and other sensors are hampered by the very high temperature at the exhaust, high content of carbon soot and frequently turbulent flow conditions, which cause a non-parabolic flow profile across the exhaust cross section. We are developing an optical imaging method for gas flux measurements, using shadow video imaging of the dynamic hot gas emission profile at the power plant exhaust. All, high exhaust gas temperature, pressure increase and carbon soot content cause small variations of the refractive index. This deflects a considerable amount direct sunlight under inclined solar illumination conditions (at approx 38 deg inclination angle), and results in a rather sharp contrast and clear shadow image of the gas flow above the exhaust. This feature is not observable in direct transmission imaging. The distant flow shadow image pattern, as seen on the plant floor, is video monitored over a short time period and the dynamic image evolution digitally processed and analyzed. The presented method is similar to the well known optical so-called Schlieren imaging technique. Initial video processing algorithms and results are presented that provide the flue gas flow velocity directly at the exhaust exit, being close to the expected values, obtained from power plant process parameters.