Turbine Inlet Temperature measurements Iva Gianinoni*, Elena Golinelli*, Sergio Musazzi*, Umberto Perini* 4th EVI-GTI International Gas Turbine Instrumentation Conference Norrköping, Svezia, 23-25 Settembre 2009 PRESENTAZIONE POWER POINT * ERSE SpA An optical sensor for gas Turbine Inlet Temperature monitoring has been developed, which is based on spectroscopic photometric measurements of the Infra Red radiation emitted in a selected wavelength band by the CO 2 molecules in the combustion gases. Aim of the sensor is to measure on line and in real time the average gas temperature on a plant section transverse to the gas flow. The developed sensor prototype is mechanically robust and thermally resistant so to withstand typical operating conditions of industrial gas turbines. To minimize installation requirements, it has been designed in such a way it can operate through a single optical access. The instrumental system consists of three units: an optical probe (i.e., a metal tube, interfaced with the combustor, cooled by water and purged with air to prevent the combustion gases from entering into the probe), a detection unit and a data acquisition unit. The IR radiation coming from the hot gases is collected and focused by a lens onto an optical fiber connecting the probe with the detection unit. Here the radiation emerging from the fiber is first collimated, then filtered by means of an IR interference filter (centered at around 4.5 µm) and finally focused on a photoconductive detector. Data acquisition and analysis is carried out by a PC based analysis unit. All the system is controlled via a Labview dedicated software The sensor validation is still ongoing both at national and European level. In the frame of the current EU funded HEATTOP research program, tests have been carried out on the Siemens HPMC rig located at the DLR Laboratories in Cologne. This facility, which is normally used for burner testing, can be conveniently selected for our purposes, since it generates gas flows at temperatures and pressures typical of an industrial gas turbine. The IR probe has been properly installed transversely to the transition piece, where temperatures as high as 1600 °C can be reached. Measurements were performed during two separate tests, where the gas flow temperature was gradually increased up to the maximum value, then maintained almost constant for a given period and finally decreased down to ambient temperature. The signal achieved from the probe was then compared with the flame temperature provided by the rig control system. As shown in Fig. 1 a good agreement between the two signals can be observed during the initial phase of the test (i.e. during heating at temperatures between around 900 and 1300 °C). Subsequently the probe signal becomes noisy and the agreement between the two sets of data significantly worsens. This is likely due to misalignments caused by heating-induced optical gradients located close to the probe tip. This problem will be overcome in the next testing campaign by proper tailoring of the probe tip. Fig. 1 – Flame temperature and detected signal versus measurement time. Flame Temperature – Signal 800 1000 1200 1400 2009/07/10 10.29.17 2009/07/10 10.32.57 2009/07/10 10.36.37 2009/07/10 10.40.16 2009/07/10 10.43.56 Time T e m p er at u r e ( ° C ) 5 15 25 35 45 S i gn al ( m V ) Flame Temperature Signal