In order to reduce dependence on fossil resources, new biomass feedstocks need to be developed to produce bio-based materials and bio-based energy. However, new biomass feedstocks can differ significantly from classic biomass
feedstocks, such as wood. These differences can cause substantial problems, especially in the conversion processes. Hence, processes suitable for new biomass feedstocks need to be adapted. A detailed characterisation of the feedstocks forms the basis for such adaptation. Within WP5, we deal with the characterisation of new biomass feed-stocks with a focus on their behaviour during different conversion processes. For this purpose, existing infrastructure is being adapted and new types of characterisation methods are being developed. Methods used start from physical and chemical analyses to gain basic data about the feedstocks. To determine conversion characteristics we start small -scale (e.g. TGA thermogravimetric analysis), moving to a larger system (e.g. bench-scale gasifiers).
We integrate the results based on experimental data and numerical analyses. One currently ongoing joint research activity is a comprehensive round robin based on TGA involving seven partners. Although TGA is widely used, the agreement of results originating from analyses with different TGA systems, often obtained by following individual protocols, is not well studied. Hence, results may differ substantially depending on the operator performing the analyses.
The aim of the round robin is to evaluate pyrolysis, torrefaction and char oxidation/gasification of new biomass feedstocks and to compare results with better known feedstocks like wood. In order to validate the TGA results against literature and to minimise potential inhomogeneity in biomass feed-stocks, Avicel Cellulose PH-105, a well-defined reference material, has been included in the round robin. In the first stage of the TGA round robin, pyrolysis of Avicel Cellulose PH-105 and beech wood (homogenised in the ISO certified lab of CERTH and distributed to all partners) was investigated. For this purpose, an experimental procedure was formulated, which was followed by each partner. The evaluation of the first stage of experiments is completed and a main result is presented with the temperature of the
maximum reaction rate (peak temperature) at a heating rate of 5 K/min. The mean peak temperature for Avicel Cellulose PH-105 was determined with 325°C, which is in good agreement with literature. However, the values of the peak temperature within the round robin ranged from 314°C – 338°C (Figure 1). A com-parable range of peak temperatures were found for the beech experiments (335°C – 364°C). The deviation only depended on the instrument and not on the material investigated. (e.g. minimum temperature for Avicel Cellulose and beech obtained by the same partner). This first data set shows that the instruments used have an influence on the results. The ongoing evaluation will help with the determination of the characteristics of the instruments and methods employed by each partner, providing a bench-mark for the experiments to follow. By identifying the differences between the various instruments and procedures, it will be possible to obtain meaningful final measurement protocols and results. In future tests within the round robin, new biomass feedstocks and different atmospheres will be investigated. Furthermore, additional partners from the European Energy Research Alliance
(EERA) will be included in the round robin in order to extend the range of TGA systems considered. The final results will allow a thorough evaluation of the comparability of TGA tests and the factors causing differences.

(thermogravimetric analysis ) experiments performed by seven partners. Each line represents a partner’s experimental data.
