Motivation: Thermochemical conversion of biomass to syngas by gasification of biological wastes like food or wood residuals wins more and more importance for energy supply because this kind of fuel is sustainable and generally available on demand. The syngas is composed of fuel components like CO, CH4, H2 and light hydrocarbons, but also contains undesired components like particulate matter and particularly tar. The latter constituents represent a complex mixture of aromatic compounds like toluene, phenol or naphthalene, which vary in relative composition and absolute concentration related to the composition of the biomass and the actual quality of the gasification process.Tar content in the raw syngas leads to complications like tar deposition on the walls, coking and clogging of pipes in the equipment of the associated processes in which syngas is used as a fuel. The continuous monitoring of the tar in the syngas even at low concentrations is, therefore, of major importance to enable installation of a feedback operation control of the gasifying process and achieve minimizing of the tar content. In the past, for analysis of tar in syngas well established, but expensive analysis methods like FTIR, NMR and GC/MS were published [1]. Also, tar analysis by LED induced fluorescence spectroscopy was reported in [2], but preliminary tests were conducted in the liquid phase at relatively high concentrations of phenol (8-10%). Recently, online monitoring of the tar concentration in producer syngas by use of a flame ionization detector (FID) was described in [3]. The FID difference signal based on the syngas stream without and with condensation of the tar (cooled filter at 20°C-100°C) represents the tar concentration. This concept of tar analysis seems to be quite accurate for tar concentrations higher than 5g/m3. New concept of more sensitive tar-monitoring in syngas: In this paper, for the first time, a novel concept of continuous tar monitoring in syngas is introduced which has been tested in a lab setup (Fig. 1a). It allows accurate measurements of toluene (used as a model tar) even below 1000ppm. The sensing principle is based on the estimation of the residual oxygen demand for tar combustion.First, a small flow (100ml/min) of the hot syngas stream is extracted from the gasifier and dosed with synthetic air to adjust stoichiometric combustion conditions (λ=1) by use of an electronic mass flow controller (MFC) operated in a feedback loop with the signal of a classical high-temperature Pt/8YSZ/Pt - oxygen concentration cell. In the second step the synthetic air flow is kept constant, but now the syngas is lead over a condensation unit (T=-32°C) and again conducted to the oxygen concentration cell (KS1D, Lamtec GmbH) or to a broadband lambda-probe (LSU 4.9, Bosch GmbH). The latter sensor provides a signal (coulometric current Ip) proportional to the excess oxygen concentration measured after tar condensation (λ>1). The difference signal DIp = Ip(λ>1) – Ip(λ=1) represents the excess oxygen concentration after tar condensation which is directly related to the oxygen demand for tar oxidation. Discussion and outlook: Of course, this method of tar monitoring does not provide an analysis of the tar content because the measured oxygen demand related to the tar concentration depends on the specific mixture of aromatic compounds forming the tar. In addition, even at temperature of approximately -32°C volatile components like toluene, which is one of the major constituents of tar, can only be condensate to a residual saturation concentration of about 800 ppm according to Clausius-Clapeyron equation, which corresponds to a concentration of about 3g/m3 (25°C). This means, at this temperature other components of tar with higher evaporation enthalpies like naphthalene, phenol will be estimated at a much lower sensitivity limit and, therefore, will represent the tar concentration even at considerably lower values than 1000ppm.In the next step this novel procedure of tar monitoring will be automated, tested with naphthalene (model tar) and finally, first experiments will be conducted with syngas of a wood gasifier. These results will be reported as well and limits of detection will be discussed in context with some technical advantages/restrains. Acknowledgement This work is part of the EBIPREP collaboration project (www.ebiprep.eu) financed by the EU International Programme INTERREG V Oberrhein 2017-2020.
Read full abstract