Detection of ethene and other hydrocarbons in gas turbine engine exhaust using non-intrusive FTIR spectroscopy

Abstract

As part of the EU funded project AEROJET2, a number of gas turbine engine tests were performed in different facilities around Europe. At Farnborough, UK a Spey engine was used to test a suite of prototype optically based instrumentation designed to measure exhaust gas emissions without using extractive probe systems. In addition to the AEROJET 2 prototype instrumentation, a Bruker Equinox 55 Fourier transform infrared (FTIR) spectrometer was used to obtain infrared spectra of the exhaust plume both in emission and absorption mode. The Bruker FTIR spectrometer was fitted with a periscope system so that different lines of sight could be monitored in the plume in a vertical plane 25 cm downstream from the nozzle exit and 20 cm upstream of the center line of sight of the AEROJET 2 prototype instrumentation. DERA (now QinetiQ) provided exhaust gas analysis data for different engine running conditions using samples extracted from the plume with an intrusive probe. The probe sampled along a horizontal plane across the centerline of the engine 45 cm downstream of the nozzle exit. The Bruker spectrometer used both InSb (indium antimonide) and MCT (mercury-cadmium-telluride) detectors to maximize the sensitivity across the IR range 600-4000 cm -1 . Typically, CO 2 and H 2 O IR signatures dominate the observed spectra of the plume. However, the engine tests showed that at low power engine conditions spectral features associated with CO around 2147 cm -1 and with hydrocarbons could be observed at around 3000 cm -1 . In particular the presence of ethene (C 2 H 2 ) was detected from observation of its characteristic in and out of plane vibration mode at 949 cm -1 . At high engine powers the presence of NO was detected at 1900.3 cm -1 . Species concentrations were calculated using a slab model for each line of sight compared against reference spectra. The engine plume was assumed to be symmetric about the centerline. On this basis, data from the extractive sampling gas analysis that had been obtained by traversing the probe across a horizontal plane through the centerline could be compared with non-intrusive measurements made by scanning vertically. Adjustments have been made to account for the 20 cm downstream offset in measurement planes of the probe and the spectrometer behind the nozzle exit.