Coal Preparation » Dewatering
Australian coal production is approximately 200Mt per annum. Most of this coal is cleaned using wet beneficiation techniques and as a consequence the overall product consignment contains entrained moisture. This entrained moisture, which normally is around 8-9wt%, presents the market with problems associated with product handling, calorific dilution and increased freight charges. Depending on the particle size of the coal being treated, it is consigned to appropriate unit operations for dewatering to meet requirements for overall moisture content.
The unit operations used for dewatering include screens, centrifuges and filters. Final dewatering may be achieved using stockpile drainage. The relative proportions of coal reporting to these operations depends strongly on the particle size distribution of the coal.
Typically, coarse coal (<35mm) is dewatered to around 4-5% using screens while the -35+0.5mm fraction is dewatered using centrifuges to around 7-8wt% and the -0.5mm is consigned to filters to provide a product of 20-22wt% moisture.
Because the largest percentage of coal is dewatered in centrifuges (~50%), the control of moisture in this size fraction is of paramount importance in the control of the overall blended product moisture.
This project has addressed this particular issue with the objective of enhancing moisture removal in this size fraction by the superimposition of a turbulent air stream within the centrifuge to reduce product moisture by a target figure of 1wt%.
The project has been carried out in three stages. The first stage demonstrated that such moisture reductions were realistically achievable at bench-scale under static conditions. The results of these investigations have been reported previously.
Following a literature survey, it was decided that the concept was sufficiently novel to warrant patent coverage which was accordingly taken out.
The present report deals with the findings of Stage 2 of the investigation which had the objective of demonstrating the effect in a continuous operating mode and to develop operating parameters which will enable design criteria to be established for the construction of a 100/t/h demonstration unit to be installed at the Catherine Hill Bay Coal Preparation Plant operated by Coal Operations Australia Lit (COAL).
Retrofitting an air purge manifold into a pilot scale vibrating basket centrifuge led to a moisture reduction of between 0.9 and 1.0wt% when treating Bayswater coal in the size range -6+0.5mm. When treating a sample of coal from Catherine Hill Bay with a top size of 35mm, the moisture reduction was between 0.6 and 0.7wt%. Possible reasons for the differences in behaviour include:
- damage to the manifold by the larger particles leading to diversion of the air stream away from the bed
- the presence of particles larger than the manifold presenting a solid rather than porous barrier to the air stream
The parameters exerting most influence on moisture reduction were the speed of the air and clearance between the manifold and the centrifuge basket. It was important to ensure the highest air speed and closest approach of the manifold to the basket without ploughing the bed of coal to achieve the greatest moisture reductions A simple air knife appears an effective design of manifold. Two identical air knives was much less effective than one for the same air flow rate, thus confirming the importance of air speed rather than simply air flow rate.
A procedure for running and sampling the pilot scale rig is now available for reliably and reproducibly demonstrating the efficiency of air purging under the appropriate test conditions.
The maximum moisture reductions achieved with the pilot scale machine (0.9-1.0wt%) were lower than those determined in bench scale tests (1.5-1.8wt%), despite using much higher air velocities. Likely reasons for this difference are:
- the mean residence time spent by any given coal particle in the air purge zone was probably below 25 milliseconds in the pilot scale machine compared with a bench scale test duration of 10 s
- in the bench scale tests, the air is introduced through a rotary bearing into the centrifuge so there was no different in angular velocity between the air and the coal bed, whereas in the pilot scale machine the air emerges from the outlet of a stationary manifold at right angles to a moving basket. Thus the air purge stream would acquire a tangential component to its velocity and hence its speed and effectiveness at penetrating the bed would be reduced
- the bench scale rig is totally enclosed so all the purging air passes through the coal bed. In the pilot scale machine not only can the air be diverted by the spinning motion of the basket, but it could simply deflect off the larger particles as they pass under the manifold rather than penetrate the pores of the bed
Bench scale tests suggested that air purging was effective at reducing moisture from a wide range of coals, but that the extent of moisture reduction increased with increasing coal rank. This relationship is reasonable since water removal would be expected to be easier for increasing surface hydrophobicity. Both coals utilised in the pilot scale trials were of low reflectance (0.7-0.8) and are therefore likely representative of the more difficult coals to dewater.
Analysis of two coals by the University of Queensland, suggested that bench scale air purging was capable of removing nearly all the surface moisture, with only internal non-centrifugable moisture (NCM i ) remaining in the product.
Initial attempts to simulate mathematically the flow of turbulent air streams through coal beds have given encouraging results. The Ergun equation gives good agreement at with experimental data at low air velocities, although deviations are more significant at high air velocities which are closer to those used in the process. More insight is needed. Eventually, the model will be used to develop predictions of pressure drop across beds of different thickness as a function of air velocity.
More pilot scale results are required, in order to:
- perform more prolonged runs than have been possible hitherto
- to assess the performance of alternative designs and orientations of manifold
- to establish the effects of other centrifuging parameters such as coal throughput, spin speed and coal size on air purge performance.
The pilot scale centrifuge rig will be relocated to the Catherine Hill Bay CPP to complete these studies.
The data from the forthcoming trials and predictions of pressure drop form the Ergun equation will be used to develop scale parameters from which air requirements and process economics can be assessed.
The bench scale results are in poor agreement with the equation developed by Firth et al. A possible reason for this rests with the lack of recognition, in the proposed equation, of the role of inherent mineral matter in the determination of three phase contact angle.
It has been agreed with ACARP that the results of on-site pilot scale trials will be reported as a supplement to this report at such time as site problems have been resolved.