A combination of image processing techniques and regression models was evaluated for predicting equivalence ratio and major species concentration (H₂, CO, CO₂ and CH₄) based on real-time image data from the luminous reaction zone in conditions and reactors relevant to biomass gasification. Two simple image pre-processing routines were tested: reduction to statistical moments and pixel binning (subsampling). Image features obtained by using these two pre-processing methods were then used as inputs for two regression algorithms: Gaussian Process Regression and Artificial Neural Networks. The methods were evaluated by using a laboratory-scale flat-flame burner and a pilot-scale entrained flow biomass gasifier. For the flat-flame burner, the root mean square error (RMSE) were on the order of the uncertainty of the experimental measurements. For the gasifier, the RMSE was approximately three times higher than the experimental uncertainty – however, the main source of the error was the quantization of the training dataset. The accuracy of the predictions was found to be sufficient for process monitoring purposes. As a feature extraction step, reduction to statistical moments proved to be superior compared to pixel binning.

Optical and intrusive measurement techniques for temperature and soot concentration in hot reacting flows were tested on a small-scale burner in fuel-rich, oxygen-enriched atmospheric flat flames produced to simulate the environment inside an entrained flow reactor. The optical techniques comprised two-color pyrometry (2C-PYR), laser extinction (LE), and tunable diode laser absorption spectroscopy (TDLAS), and the intrusive methods included fine-wire thermocouple thermometry (TC) and electrical low pressure impactor (ELPI) particle analysis. Vertical profiles of temperature and soot concentration were recorded in flames with different equivalence and O₂/N₂ ratios. The 2C-PYR and LE data were derived assuming mature soot. Gas temperatures up to 2200 K and soot concentrations up to 3 ppmv were measured. Close to the burner surface, the temperatures obtained with the pyrometer were up to 300 K higher than those measured by TDLAS. Further away from the burner, the difference was within 100 K. The TC-derived temperatures were within 100 K from the TDLAS results for most of the flames. At high signal-to-noise ratio and in flame regions with mature soot, the temperatures measured by 2C-PYR and TDLAS were similar. The soot concentrations determined with 2C-PYR were close to those obtained with LE but lower than the ELPI results. It is concluded that the three optical techniques have good potential for process control applications in combustion and gasification processes. 2C-PYR offers simpler installation and 2D imaging, whereas TDLAS and LE provide better accuracy and dynamic range without calibration procedures.

In the present work, 5 different axisymmetric burners with different directions of the oxidizer inlets were experimentally tested during oxygen blown gasification of torrefied wood powder. The burners were evaluated under two different O2/fuel ratios at a thermal power of 135 kWth, based on the heating value of torrefied wood powder. The evaluation was based on both conventional methods such as gas chromatography measurements and thermocouples and in-situ measurements using Tunable Diode Laser Absorption Spectroscopy. It was shown that changes in the near burner region influence the process efficiency significantly. Changing the injection angle of the oxidizer stream to form a converging oxidizer jet increased process efficiency by 20%. Besides increased process efficiency, it was shown that improvements in burner design also influence carbon conversion and hydrocarbon production. The burner with the best performance also produced less CH4 and achieved the highest carbon conversion. The effect of generating swirl via rotating the oxidizer jet axes was also investigated. Swirl broadened or removed the impingement area between the fuel and oxidizer jets, however resulting in differences in performance within the measurement uncertainty. 

The work demonstrates the performance of the optical extinction technique for real-time diagnostics of the fluctuations in biomass particle flows. The online measurements of fluctuations of density were used to determine the biomass particle mass flow fluctuations. Biomass flows were produced using laboratory biomass particle feeder (mass flux up to 10 g/min) and the hopper-screw feeding system of the pilot-scale entrained flow rector, mass flux up to 500 g/min, located at SP ETC in Piteå. The experiments showed that the time-averaged extinction appeared to be linearly related to the real particle mass flow. The relatively fast variations in biomass feeding rates measured using the extinction technique were confirmed by fast balance measurements (in laboratory feeder experiments) and by real-time tunable diode laser CO and H₂O concentrations measured in the reactor core of the entrained flow gasifier.

This paper reports on the development of the tunable diode laser absorption spectroscopy sensor near 4350 cm−1 (2298 nm) for measurements of CO and H2O mole fractions and soot volume fraction under gasification conditions. Due to careful selection of the molecular transitions [CO (υ″ = 0 → υ′ = 2) R34–R36 and H2O at 4349.337 cm−1], a very weak (negligible) sensitivity of the measured species mole fractions to the temperature distribution inside the high-temperature zone (1000 K < T < 1900 K) of the gasification process is achieved. The selected transitions are covered by the tuning range of single diode laser. The CO and H2O concentrations measured in flat flames generally agree better than 10 % with the results of 1-D flame simulations. Calibration-free absorption measurements of studied species in the reactor core of atmospheric pilot-scale entrained-flow gasifier operated at 0.1 MW power are reported. Soot concentration is determined from the measured broadband transmittance. The estimated uncertainties in the reactor core CO and H2O measurements are 15 and 20 %, respectively. The reactor core average path CO mole fractions are in quantitative agreement with the µGC CO concentrations sampled at the gasifier output.