Coal Preparation » General
The objective of this project is the identification of the factors controlling the efficiency of the high capacity Banana screens when they are used to recover media via drain and rinse. This includes potential approaches to improve efficiency with respect to fluid recovery, partition curves and magnetite recovery.
Previous ACARP project (C7048 "Performance of Banana Screens in Drain and Rinse Applications") was focussed on investigating the performance of banana screens in a number of operating plants. The results from this audit identified that the efficiency with respect to fluid recovery, partition curves and magnetite recovery were linked to; the volumetric flow of the feed to the screen, the solids content of the feed slurry and the effective aperture and open area of the screen. These results provided a base line for the experimental conditions to be used in this project.
While ACARP project C7048 identified the three main factors that effect screen efficiency this project concentrated on varying these factors and others on a well controlled pilot banana screen at CSIRO's Energy Technology pilot plant located at the Queensland Centre for Advanced Technologies at Pullenvale in Brisbane's west.
Variables tested were panel 1 (P1) angle (angle from the horizontal of the first of two panels on the pilot screen), feed flow rate, feed solids content, G force (stroke), frequency of vibration, aperture size and open area, screen motion (linear or rotary) and panel type (manufacturer). The data collected from these experiments was used to produce empirical and fundamental models for banana screen operation.
The change in initial slope angle (P1) at 50% solids feed had little effect on the drain rates of panel types B and C, however panel type A showed some significant variation in drain rates with changing P1 angle with the highest drain rate recorded for a P1 angle of 25 degrees. At higher solids concentrations significant variations of drain rate with changing P1 angle were observed when compared to the lower solids content feeds.
Increases in volumetric flow rate decreases the overall drain rate regardless of solids content.
Solids concentration has a major effect on drain rates and partition curves. At the higher solids content used in these experiments (ca. 60%) drain rates were reduced by almost 50% when compared to the lower solids content feed (ca. 50%) and are more sensitive to P1 angle changes.
The drain rates remained relatively constant over a range of vibration forces, from approximately 2G to 4G. Increasing the vibration force above 4G results in a reduction in drain rate. This implies that it may be possible to operate at a lower vibration force while maintaining screen efficiency, thus reducing mechanical stress and wear.
Linear motion produces marginally better drain rates then rotary motion. However, individual drain rates for panels P1 and P2 indicate that rotary motion results in a lower drain rate on the lower angle deck then linear and higher drain rates on steeper inclined panels (eg. P1 35 degrees).
At lower G forces the frequency of vibration has a minor effect with maximum drain rate occurring from approximately 13 to 17 Hz. At the higher G force (7.1G) however, a marked decrease in drain rate can be observed from 13 Hz and higher.
Two different feeds were used to test the effect of magnetite content, and were found to have little effect on the total drain rates. Feed 1 contained 45% w/w magnetite and Feed 2 33% w/w.
As expected, increasing the panel aperture (and thus open area) increases the drain rate. The 1.0mm aperture (new) had drain rates of over 90% at lower feed rates and over 80% at the highest feed flow rate. The 100% worn panel with a nominal aperture of 1.03mm was expected to be close to the drain rates of the new 1.0 mm panel, but the measured drain rates were closer to the 50% worn and new 0.5 mm panels than the 1.0 mm new panel.
The partition curves show a marked difference between the nominal 1.0 mm and 0.5 mm panels. The 1.0 mm panel's partition curve was steeper with a lower Rf and higher S50 then that of the 0.5 mm panel. This leads to more efficient medium recovery with a drier oversize. Both curves have S50's approximately 0.1 mm below the nominal panel aperture.
In all cases higher drain rates were recorded for panel Type A then for the other two panels. Type A was a non-continuous slot design with softer polyurethane then Types B and C. The softer material resisted blinding and pegging to a greater extent than the other two screen types leading to better drain rates and a lower Rf. The degree to which the screens will wear was outside the scope of this project and no comment can be made on the wear characteristics of the softer material.