Modelling of ash melts in gasification of biomass

Abstract

The need for advanced biofuels produced from sustainable sources is stressed, both on national and international level due to a global agreement to limit the Earth’s global warming. The major goals in the Norwegian agreement on climate policy are to become climate neutral by 2030 and to become a net-zero emission society by 2050. One of the priority areas for action is to reduce the sources of greenhouse gases by speeding up the introduction of low-emission alternative transport fuels, such as liquid transport biofuels.

A well-known process for converting biomass resources into liquid transport biofuels involves gasification, a thermochemical process that converts the biomass into a gaseous mixture of syngas in the presence of heat and a gasifying agent. The syngas consists of mainly hydrogen (H2) and carbon monoxide (CO), and can be further processed into biofuels. Among the different technologies applied for biomass gasification, fluidized beds have industrial advantages due to the ability to process a wide range of biomass under controlled operating conditions. The fluidized bed gasifiers also offer several other advantages, including good mixing, high heat and mass transfer and high productivity at a relatively low process temperature. However, processing biomass-derived fuels in fluidized beds suffers from ash related problems. The major challenge is associated with molten biomass ash and the formation of agglomerates that cause fluid dynamic disturbances in the bed. If not counteracted, the bed disturbances lead to operational problems that might result in decreased efficiency, high maintenance costs and unscheduled shutdowns. Bed agglomeration and de-fluidization are closely linked to the ash melting behaviour, and has been reported as one of the problematic issue prohibiting an economical and trouble-free operation. Hence, the key to unlocking fluidized bed biomass gasification as a viable route for biofuels production is by solving the challenges related to the ash.

This PhD thesis addresses the key issues related to bed agglomeration and de-fluidization in fluidized bed gasifiers. Experimental work and computational modelling were combined in order to achieve a fundamental understanding, and insight into the underlying mechanisms of the ash melting behavior and the bed agglomeration processes. The main objective was to develop effective and accurate methods and models to be used in prediction of the agglomeration tendency of different types of biomass during gasification in fluidized beds. The overall approach was divided into three sections: (i) CPFD simulations combined with fluidization and gasification experiments to gain necessary knowledge on the fluidization characteristics, (ii) fluidization experiments to generate new sets of data that could form the basis for (iii) a mathematical model for prediction of the critical amount of accumulated ash/bed material in the gasifier. The experiments were carried out in three different fluidized bed systems: (i) a cold flow model, (ii) a 20 kW laboratory scale model, and (iii) a micro-scale model. The commercial CPFD software package Barracuda Virtual Reactor was used for the computational part. The investigated biomass samples were grass, wood, straw and bark.

The results point out that the operating temperature and the composition of the major ash forming, in particular Si, K and Ca, are significant factors leading to ash melting problems in fluidized bed processes. Additionally, the findings show that the ratios between the major ash forming elements, K, Si and Ca, in the biomass play an important role in the agglomeration process, and that different combination of those elements are especially problematic when processing biomass fuels in fluidized bed systems. The results also indicate that bark tended to have the highest tolerance limit of accumulated ash in the bed for all the investigated temperatures. For example, the ash/bed material was measured to 7% by weight at 900°C, compared to grass (3%), straw (1%) and wood (1%).

A multiple regression was calculated to predict the mass ratio of accumulated ash/bed material based on the operation temperature (T) and the mass ratios of (Si/K) and (K/Ca). The final model expresses the amount of accumulated ash/bed material at the onset of bed agglomeration and de-fluidization:

Accumulated ash/bed material (wt %) = 17.06 – 0.02·T + 4.04·(Si/K) + 1.05·(K/Ca) The overall regression was statistically significant (R2 = 0.81, F (3, 30) = 38, p<0.0001).