A study of the uncertainty associated with tar measurement and an investigation of tar evolution and composition during the air-blown fluidised bed gasification of torrefied and non-torrefied grassy biomass

Miscanthus×giganteus (M×G), a typical bioenergy crop, has been widely planted in Ireland. But the virgin miscanthus is considered as a low grade fuel. The aim of this study is to investigate a method to (a) investigate the stage of tar analysis methods and implement one of them; (b) to upgrade the fuel properties of raw miscanthus via torrefaction and to compare tar evolution over a set of process conditions during the bubbling fluidized bed gasification; (c) assessment of the measurement uncertainty of solid phase adsorption–gas chromatography (SPA-GC) measurement system in order to enhance the reliability of the reported results. Overview on tar measurement methods has been conducted in order to understand the needs and challenges of gasification developers in association with the tar characterization. Tar measurement methods are divided in off-line and on-line methods. Off-line methods are based on trapping the tar by condensation on cold surfaces or filters, by absorption in a cold organic solvent or by adsorption onto suitable sorbents. Subsequent analysis is mostly performed by GC or gravimetrically. On-line characterize tars in the hot gas phase and offer quick real time information. However, in practice, on-line monitoring of tar content in the producer gas is not performed. Off-line tar measurements are conducted only occasionally by research institutes in the context of research projects. The main part of research involved comparing the gasification of raw and torrefied M×G. The experiments were conducted in an allothermal air-blown BFB gasifier using olivine as bed material at ECN in Netherlands. There is an indication that torrefied M×G generates higher yields of both total GC detectable tars as well as 20 individually quantified tar species. The total GC detectable tar production from raw M×G is 14 to 19 gTotal GC detectable tar kg-1 Biomass-daf whereas tar yield for torrefied M×G varies from 21 to 31 gTotal GC detectable tar kg-1 Biomass-daf measured for the temperature range from 715 to 850 °C and from 660 to 850 °C for raw and torrefied M×G, respectively. High tar content is attributed mainly to the higher lignin, but lower moisture content and the small particle size of the torrefied M×G may also contribute. In terms of the effect of the operational conditions, temperature is demonstrated to play the dominant role in tar reforming reactions namely (a) tar cracking and (b) PAH enlarging. The highest yields of total GC detectable tar, secondary tars, and tertiary-alkyl tars are typically observed between 700 and 800 °C, while tertiary polyaromatic tars increase up to the tested maximum temperature of 850 °C. The data from the two experimental campaigns suggests that at constant temperature the ER has relatively little impact on the amount or composition of tar. For the SPA-GC measurement system used for tar sampling and analysis the uncertainty analysis were performed. The two chromatographic methods studied do not give comparable total gas chromatography (GC)-detectable tar results in the product gas, although a trend of decreasing tar yields across gasification temperature range was observed for both sets of measurements. The GC-FID measurements were significantly higher than the GC-MSD measurements. Their overall uncertainties also vary by a significant margin. While a quantitative method based on a single calibration curve offers a significant advantage, in terms of speed and simple quantitation of total GC-detectable tar, such an approach introduces greater uncertainty within the reported results. The relative expanded uncertainty is 109.4 % for the GC-MSD based measurement system and 35.0 % for the GC-FID based measurement system. The dominant uncertainty contributor arises from the chromatographic category and the related quantitation method. Overall, the results reveal that the torrefaction of M×G is not a promising approach for tar reduction in BFB gasification. Tar measurement systems and their calibration methods are the elements of unreliability of the results. As such, they remain the on-going challenge for gasification developers.