Impact of Milling Techniques on Coke Oven Feed

The international export market for Australian coals is dominated by coking coal sales. However, few Australian coking coal producers process their coal domestically to produce direct coke oven feed. Therefore there is little understanding of how best to prepare each coal to achieve optimum coke quality and battery operations at the user end.

Coke quality is controlled by coal properties (rank, ash yield and composition), bulk material properties (size distribution, bulk density and moisture content) and coke oven parameters (oven type, heating rate, coking time, consistency of operation, method of oven loading and quenching). It is also controlled by feed preparation, in particular the best grinding process to achieve an optimum size distribution for coking behaviour.

Currently, coke oven feed is crushed to size specifications of approximately 85% passing 3mm. It is known that some coals require more energy than others to meet this specification, and generally, adjustments to milling configurations are made empirically. The major objective of this project was to develop models for the swing hammer mills under varying machine conditions to assist in the beneficiation of coke oven feeds. It was required that the models be capable of predicting hammer mill products for given feed properties and machine configurations. Alternatively, to achieve a required product from a given feed the models should be able to indicate the machine settings. The outcome of a successful simulation model is the capability to simulate the impact of changing blends or mill configurations to achieve a desired coke oven feed size distribution.

This project combined studies by JKMRC, Callcott Consulting Pty Ltd (CCPL), the CSIRO and BHP Port Kembla Steelworks. The approach of the project was to:

• Conduct an audit on the industrial hammer mill operation at Port Kembla and collect feed samples for the BJD hammer mill tests;
  
• Characterise the laboratory breakage behaviour of coal using the single particle drop weight test developed at the JKMRC. Tests on coal samples from individual lithotypes and the Port Kembla (PK) hammer mill feed were used to generate the breakage functions for JKMRC modelling;
  
• Conduct pilot scale hammer mill tests at varying operating conditions;
  
• Use the data to develop and calibrate hammer mill models for both the pilot scale BJD and the industrial mill at Port Kembla.

Two models were developed and are presented in this report. One uses the JKMRC approach to energy-based mechanistic modelling (Part I), the other the Callcott mathematical model with a fundamental understanding of hammer mill behaviour (Part II). Spreadsheet versions of both models are available from the authors (Frank Shi, JKMRC and Tom Callcott, Callcott Consulting Pty Ltd).

Joan Esterle - phone 07 3327 4411, joan.esterle@csiro.au
Tom Callcott - phone 02 4968 3408, consulting@callcott.com.au