Combustion of biomass is a promising approach to reduce greenhouse gas emissions, particularly when combined with carbon capture and storage. While developing future combustors and retrofitting existing infrastructure both require a profound understanding of the underlying processes, comprehensive experimental investigations are hindered by the limited accessibility of industrial combustors, preventing the application of advanced diagnostics.

A new paper out of the Technical University of Darmstadt demonstrates that spatially resolved information on the gas phase composition (H₂O, CO₂, CO, CH₄) and temperature can be inferred from single line-of-sight tunable diode laser absorption spectroscopy (TDLAS) measurements, circumventing the need for extensive optical access. Inferring spatially resolved information in the gas phase from path-integrated transmission spectra renders an ill-posed inverse problem, which is solved in a Bayesian approach that incorporates available prior information.

The results obtained from applying the technique to a semi-industrial 1MW_th combustion chamber are found to compare well to those of an extractive gas sampling analyser. The developed method enables a detailed characterisation of the combustion process and offers insights into the effects of flue gas-recirculation. It enables comprehensive investigations of industrial systems with minimal optical access. In addition, it has the potential to eliminate the time-consuming traversing and calibration required by extractive systems and to be used for routine monitoring.