Development of a tar decomposition model for application in a Chemical-Looping Reformer operated with raw gas from a biomass gasifier

The production of Synthetic Natural Gas (SNG) represents one of the promising alternatives for biofuel manufacture. The transport sector is where SNG has been identified as having the highest potential in terms of profitability and use efficiency, making it the main aim of production. The key process steps to yield SNG are thermal gasification of biomass followed by methanation of the product gas. But, before reaching methanation the producer gas has to be cleaned from the presence of organic hydrocarbons called tars as well as other contaminants, eliminating them from the mixture of permanent gases. Tars are usually referred to as condensable hydrocarbons that start to condense already at temperatures around 350°C. As the tar condenses it creates operating problems like clogging and blockage of equipment downstream the gasifier. A system for cleaning the producer gas from biomass gasification was developed at Chalmers University of Technology, using a Chemical-Looping Reformer (CLR) for catalytic cracking of tar components. The system was developed for further implementation in the industry with the aim of making it a quicker solution for tar cleaning. The objective of the work was to develop a model for catalytic decomposition of tars, using data available by experiments occurring within the CLR-System at Chalmers. The system is fed with producer gas from Chalmers 2-4MWth biomass gasifier, which goes into the dual-fluidized bed process in which the system consists. The available data for the work was achieved by running the CLR-System with a manganese-based catalyst in the fuel reactor (FR) at three different working temperatures and two oxygen concentrations for reforming of the catalyst in the air reactor (AR). Development of the decomposition model was done firstly by grouping the analyzed tar molecules according to structures, conversions and amount and secondly by study of models describing decomposition processes. From implementation of the developed model ruling first order differential equations in the mathematical software MatLab, it was possible to verify to what extent the model correctly describes the decomposition processes inside the reactor and to have a first impression on how fast the reactions are or how each reaction interacts with the others. Experimental data and simulation results only differed by around 15% maximum. It was conclusive that increasing temperatures and higher oxygen concentrations perform better than lower values, but can also have an influence on the composition of the permanent gases. It was possible to detect some trends on the decomposition pattern but no correlation between working conditions and cracking processes can be made. Finally, temperature seems to have a higher influence on the results than oxygen concentration