Industrial Study of a Novel Dense Medium Cyclone at Gregory CHPP

Following the development of a novel dense medium cyclone at a pilot scale at the JKMRC, a 1000mm diameter large dense medium cyclone was designed and fabricated. The cyclone, called JKDMC, was tested at the Gregory Coal Handling and Preparation Plant over the period between 28 July and 29 August 2003.

An existing 1000mm DSM cyclone was replaced by the JKDMC, and was operated in parallel with another DSM cyclone in the same module. The two DMCs had similar inlet and O/F diameters, but a different U/F diameter.

The cyclone performance was assessed by two methods: density tracer tests and sampling. 10 density tracer tests were conducted using 32mm cubic tracers with narrow increments (0.01 RD). Tracer number for each density was 10 and 30 respectively, and the latter number delivered more accurate results. The Ep value was 0.006-0.008 for both DMCs.

Three sampling tests were carried out, two with the JKDMC using different vortex finders, and one with the DSM cyclone. In order to operate on a similar medium-to-coal ratio, as defined by the plant personnel, the feed pressure used for the tests was 80 kPa for the JKDMC and 100 kPa for the DSM cyclone. The pressure difference makes the performance assessment more difficult.

32mm tracer tests showed no difference between the two units. However, sampling tests showed that the JKDMC performance was poorer than that of the DSM cyclone on every size fraction. The Ep value on the entire size fraction (50-0.5mm) was 0.022 for the DSM and 0.043 and 0.035 for the JKDMC.

The major reason for the poor separation in the JKDMC was the lower feed pressure. The feed pressure of 80 kPa is just equivalent to 6.37D where D is the cyclone diameter, leading to a low density differential. Unfortunately, it was not possible to test the JKMRC at the required feed pressure of 9D at Gregory, due to insufficient pump capacity. Another possible reason for the poor performance was the inlet design of the JKDMC, which was different from the successful pilot one. The original consideration for the design was to ensure a large feed rate for the industry-sized cyclone; however, such a design may affect to some extent the centrifugal force.