High temperature in-situ IR laser absorption CO-sensor for combustion control

In view of improved control of the combustion process in e.g. industrial glass furnaces, accurate in-situ COgas sensors are required. With help of reliable in-situ gas sensors, burner control can be optimised. By better burner control, NOx emissions can be reduced and the radiative heat transfer from combustion chamber to the glass melt and thus the energy efficiency of glass furnaces can be improved. Currently, for continuous monitoring of the combustion process of glass melting furnaces, solid state oxygen (ZrO2 /Pt) sensors are used, occasionally. In order to avoid post-combustion in the exhaust of glass furnaces, there should be sufficient oxidant, air or oxygen, to assure complete combustion. However, large air excesses lead to reduction in the thermal efficiency of the furnaces and promote NOx formation. The residual oxygen concentration in the exhaust gases partly indicate the completeness of the combustion process. However, especially at nearor under-stoichiometric combustion conditions, oxygen measurement in the flue gasses downstream is less suitable for accurate burner control. In-situ CO detection in burner ports at 1400 – 1500 °C, when possible in combination with NOx, is a better and more sensitive method for optimisation of the combustion process. For this purpose the following measuring techniques can be considered: • solid state CO-specific sensors; • on-line UV emission spectroscopy; • IR diode laser absorption spectroscopy No CO-specific solid state sensors are available which can withstand the high temperatures near the flames of about 1400 – 1500 °C. UV emission spectroscopy of flames offers the possibility to detect the presence of radicals like OH and CH in the flames, however the correlation to COand NOx-levels in the combustion atmosphere is uncertain. The principle of continuous CO-detection with IR laser-absorption techniques in hot flue gasses has been demonstrated in a feasibility study, carried out by TNO in the past. A major obstacle for practical implementation of the technique, developed in this study, was the technically complex and very expensive IR equipment (nitrogen-cooled laser and detector equipment between 4.5 and 5 μm). Therefore, the possibility for CO-detection at other (weaker) absorption lines of CO in the near infra red (between 1.55 and 1.6 μm) is under study now. In this spectral region relatively cheap optical components are available, which are used for optical telecommunication (diode lasers, fibre optic components, detectors). Using the well-known HITRAN database, several CO-absorption bands between 1.55 and 1.6 μm have been identified. Currently, the attainable accuracy of this method and the interference with other high temperature absorption bands (of H2O and CO2) are being examined. In the paper recent experimental results will be presented.