Predicting Dilatation Of Coal Blends From Models Of Softening, Bubble Growth And Gas Transport

The objectives of this project were to

· Modify/develop existing models of bubble growth and collapse within a viscous fluid, gas evolution and gas flow in porous media that represent the conditions prevailing within the plastic phase of coal;

· Use the developed model to replicate conditions in the dilatometer process, compare results with actual dilatometer measurements and explain the dilatation behaviour;

· Use the model as a predictive tool for coal blends.

This report will demonstrate that the objectives have been achieved. The project was conducted in a series of steps.

Firstly, a model of the dilatation process was developed based upon component models of gas evolution, particle swelling (due to bubble growth) and gas transport (out of particles, out of the dilatometer tube as well as into bubbles).

The developed model was then used to "fit" key parameters to a range of dilatometer experiments on single coals. The parameters represent the number of initial bubble sites in the coal, the fraction of coal particles that remain unsoftened and the inherent permeability of the sample. These quantities all have meaning for the specific coal and are shown to be consistent with other information.

Finally, the model was used to compare predictions of the dilatation of a range of binary, ternary and multi-coal blends with experimental values. In most cases, the comparison was very good, but in cases where "non-expanding" coal dominates the blend, the model has difficulty reproducing the measured dilatation curve. There is a need to obtain more dilatometer test results to strengthen the ability of the model in this regard.

The successful comparison leads us to conclude that we now have a model that is able to make reasonable predictions of the dilatation of coal blends, once a few basic parameters of the coals comprising the blend are known.

It is important to understand that the modelling process used here is based on models of the various phenomena occurring during the heating of coal particles. As a result, the model is more than a predictive tool for dilatation, since it helps considerably in our understanding of the process of conversion of coal into coke. Moreover, there is a compelling argument to extend the application of this work to other methods of testing coking coals. In particular, we see a clear development path for the model to next consider sole-heated oven/Sapoznikov tests. The key difference in those tests from the dilatometer is the presence of a vertical temperature gradient in the sample, which means that particulate coal, plastic phase, semi-coke and coke may all exist simultaneously. Ultimately, success in modelling at that level will lead to modelling of the coke oven charge itself.