Factors Affecting Coke Size and Fissuring During Cokemaking Part 2

This work addressed the mechanism of fissuring during metallurgical cokemaking and extends on a previous ACARP project. Considerable progress has been made in understanding the mechanism by an integrated modelling work / experimental program. The work points to opportunities for controlling the carbonisation process to achieve required outcomes in terms of coke lump size and distribution.

Key outcomes of the work were:

• A methodology for determination of the minimum spacing between fissures, and the time at which period doublings (i.e. every second fissure stopping) occur, has been developed and implemented.
• Implementation of the methodology allows the determination of the complete fissure pattern (at least those perpendicular to the oven walls) for a given coke oven/charge configuration.
• A plausible mechanistic explanation of the formation of lateral fissures (those that are parallel to the oven walls) has been developed.
• Knowledge of the fissure pattern, combined with the lateral fissuring explanation, enables the determination of typical size and shape of lumps after stabilisation of the coke.
• A range of modelling experiments to examine the effect of heating rate, shrinkage and coke breakage strength has been performed.  As expected, increasing heating rate decreases the fissure spacing, as does increased shrinkage or decreased strength.  But the shapes of the resultant lumps is controlled by the lengths of the stopped fissures, and heating rate, shrinkage and coke breakage strength all have different effects on this behaviour.
• In all scenarios examined, either 2 or 3 period doublings were predicted.  It is expected that 3 period doubling is likely to lead to a greater spread in lump sizes than 2 period doublings.
• 3-dimensional modelling was developed as part of the project, but it is extremely difficult and time consuming, without apparently adding much to the understanding.  Accordingly, all of the above developments are based on 2-dimensional modelling.  
• The results point to opportunities for controlling the size and distribution of lumps by means of heating or blending of the charge.
• Experimental studies compared the shape factor of coke prepared in a pilot oven with model expectations for two different heating rates and for coals of different Gieseler solidification temperature (rank). Satisfactory agreement was found. Other aspects of the modelling were evaluated by comparison with three dimensional CT scans of partially coked coal charges.

• Methods for improving the measurement of fundamental properties required by the model were examined.

• Attempts were made to improve the measurement of contraction rate of semi-coke with limited success.
• Methods were examined for dynamically measuring the Young's modulus during the heating of coal. A proposed method was developed in conjunction with the University of Queensland. It is untested.
• the energy required to created new fracture surface area remains difficult to measure.

• Simulated fissure patterns were used to generate “theoretical size distributions”, for comparison with real size distributions, by

• including a probabilistic model of breakage;
• taking account of variability in lump dimensions in real charges;
• considering a model of progressive breakage to simulate stabilisation;
• comparison of theoretical size distributions with real size distributions. This indicated that coke was often not “fully stabilised”.

• The theoretical size distribution results were used to assist improvements to the statistical modelling of coke size distribution, particularly with regard to the spread of the distribution. In addition it was demonstrated that the Brown-Wohletz distribution gave a better description of the coke size distribution than the Rosin Rammler. It remains difficult to accurately predict the spread of the size distribution.