The Queensland coal industry produced 154 million tonnes of coal in 2002-03, with 129 Mt for export. The Port of Hay Point consisting of the Hay Point Services Terminal and the Dalrymple Bay Coal Terminal (DBCT) collectively exported 78 Mt of coal in 2004, making it the world’s largest coal export port. The conveyor belts carry around 600 tonnes/hour of coal, so that the cost of the study equated to 69 sec of normal operation.
In order to meet market demand there is often great pressure on handling facilities to increase throughput, and one way of doing this is by increasing efficiency. This project focused on reducing inefficiencies associated with one of the critical elements of the handling system, the transfer chute. The transfer chute at DBCT was used as a model for the study. Figure 1 shows the transfer station at DBCT and Figure 2 shows a schematic representation of the model. To give a sense of scale, this machinery is 10m across.
|
|
|
|
Figure 1: A critical transfer station at DBCT which was modelled using discrete element analysis and HPC. |
Figure 2: Schematic representation of the transfer chute model. |
Design problems in the transfer chute can lead to spillages and blockages, which quite often require extensive manual clearing, with significant losses of production time. Velocity mismatches between the coal exiting the chute and the receiving conveyor belts removing the coal can lead to operational problems in the receiving belts, while the high speed of the coal in the chute can lead to excessive chute wear.
This system was modelled using discrete element modelling (DEM), a computational technique that treats granular material as a collection of 2D or 3D rigid particles. DEM uses properties of the coal such as stiffness, damping and friction coefficients and cohesion/adhesion levels, as well as boundary conditions and the geometry of the transfer chute, and then calculates the forces on individual elements due to contact with walls and other elements of the flow. The flow is then modelled using explicit time integration schemes to solve the equations of motion.
Given that in this particular situation the particles range in size from microns up to around 50mm, and the scale of the apparatus can be up to 10m, there can be literally billions of particles passing through the system at any point in time. This means that modelling the flow is an incredibly complex process, requiring the use of high performance computing (HPC) facilities.
Effects of perturbations in the geometry of the transfer chute were studied in order to maximise throughput and minimize the possibility of blockages occurring. Changes in the chute wall friction were modelled to examine the effect on flow velocity. Typical results used in analysis are shown in Figure 3.
|
|
|
Figure 3: Diagram showing velocities (in mm/s) of particles in transfer chute. |
Results of the analysis were presented directly to industry through on-site presentations and workshops, allowing engineers to see the predictive power of DEM so that it may be used in future chute design.
This project was a collaboration between Jeff Loughran of JCU, Peter Jacobs of UQ and engineers from Babcock & Brown Infrastructure.
Participants
Prof Jeff Loughran
School of Engineering, JCU
Dr Peter Jacobs
School of Engineering, UQ
Industry Participant
Mr Peter Wotherspoon
Engineering Manager (DBCT expansion),
Babcock & Brown Infrastructure
(previously Prime Infrastructure)
Reports:
Final Report - December 2005
(766 KB PDF)
Progress Report - September 2005
(2.5 MB PDF)
Project Proposal - April 2005
(280 KB PDF)