Industrial Trials of Novel Cyclones

Industrial trials of two JK cyclones – a 1000mm large JK dense medium cyclone (JKDMC) and a 375mm JK classifying cyclone (JKCC) – were conducted at Dartbrook CHPP and Integra CHPP respectively in the second half of 2006. They were compared with a 1000mm Multotec DMC (M-DMC) and a 375mm Krebs CC under various operating conditions. The industrial studies have improved the understanding of the relationship between the JK cyclone performance and operating conditions, in particular medium flowrates and feed pressures.

JKDMC Study at Dartbrook CHPP

The coal washed at Dartbrook during the test period was low ash coal with product yield as high as 93 to 96%. In the first series of tests (Tests 1-4), the raw coal ash content was about 9%, and in the second series of tests (Tests 5-7), it was 12%. Feed rate to the DMCs was around 220 tph and 170 tph for the first and second series of tests respectively.

The Dartbrook CHPP has a single DMC circuit and the M-DMC and JKDMC were tested in turn in the circuit using the same feed pump. Seven sampling tests were carried out comprising 5 tests for the JKDMC and two tests for the M-DMC. The Coal flowrate of DMC feed, product and rejects was specially measured for every test.

For the same DMC feed pump output or pump speed, the feed pressure in the JKDMC was 9 to 26% higher than that in the M-DMC, while the feed medium flowrate in the JKDMC was about 14% lower than that in the M-DMC. Correspondingly, operating head in the JKDMC was 20% to 31% higher than that in the M-DMC.

It has been demonstrated that the DMC performance is closely related to the feed medium flowrate or pump speed. In the current study, the JKDMC was better operated at lower medium flowrate (~788 m3/h) and pump speed (~375 rpm). On the other hand, the M-DMC was better operated at higher medium flowrate (~994 m3/h) and pump speed (~402 rpm).

Within the 7 tests, the best separation performance was achieved by the JKDMC Test-2 in the first series of tests and the M-DMC Test-5 in the second series of tests. Ep value for the size fraction 11.0-1.4mm and the shape of partition curves was very similar. However, the performance achieved by the JKDMC in Test-2 occurred under tough operating conditions – at a 28% higher coal feed rate, a 21% lower feed medium flowrate and a 38% lower medium-to-coal ratio in comparison to the M-DMC Test-5.

The advantages of the JKDMC are:

• It can be efficiently operated at a high coal feed rate, a relatively low medium feed rate and medium-to-coal ratio if a certain operating head is used.  Consequently, the magnetite consumption in the medium circuit of a plant would be reduced.
• Offset of the JKDMC is normally higher than that of the M-DMC. A required cut point can be achieved at a lower feed medium density using the JKDMC leading to less consumption of the magnetite.
• It can have higher feed pressure or operating head and lower pump current in comparison to the M-DMC at a same speed of the DMC feed pump.

In summary, the test results at Dartbrook have revealed that the performance of the JKDMC was similar to that of the M-DMC, but at a relatively higher throughput and lower medium flowrate. Magnetite consumption is able to be reduced if the JKDMC is applied.

JKCC Study at Integra CHPP

The JKCC study was conducted in a Krebs CC cluster where 9 out of 12 cyclones are normally operated. A 375mm Krebs CC at the No.6 position in the cluster was replaced by the 375mm JKCC, and its performance was compared with a Krebs CC at the No.2 position. A total of 14 tests were carried out, and the feed pressure of the cluster was progressively increased from 75 kPa to 120 kPa. The feed % solids were in the range 11.7-15.1%.

A spigot blockage problem was encountered with the JKCC when the feed pressure was 100 kPa or above. This represents an over high % solids in the U/F caused by very strong centrifugal force in the cyclone at a high feed pressure. Such a problem can be solved by adjusting spigot diameter or modifying some cyclone parts.

Separation performance of both the CCs was poorer than expected. The reason is not exactly known, and it might be caused by too high feed rate (overloading) and/or turbulence in the feed distributor of the cluster.

The best performance for both the CCs was achieved at a low feed pressure of 75 kPa, and the separation criterion α value was 2.299 and 1.600 for the JKCC and M-DMC respectively (the higher α value indicating a sharper cut).

The performance of both the CCs became poorer as the feed pressure was increased to 110 kPa or above. It is probably due to too high feed rate at high feed pressures. The nomination of 130 kPa feed pressure to the cluster in plant operation is questionable, and the feed pressure should be lowered. Alternatively, the number of Krebs CC operated in the cluster should be increased.

Based on efficiency curves of the CC tests, the JKCC performance was generally better than the Krebs CC at different feed pressures except for the tests at the 100 and 110 kPa feed pressures in which the JKCC spigot was partially blocked.

For the Krebs CC, the corrected efficiency curves were really flat, and a considerable amount of fine particles reported to the U/F although few coarse particles appeared in the O/F. The cut point was very low, in the range 0.06-0.08mm. Consequently, the efficiency of the downstream processes (Teeter beds and spirals) would be significantly affected. For the JKCC, a better U/F with a low water recovery was produced and its cut point was at 0.10-0.12mm.

The advantages of the JKCC are:

• It has better separation performance than the Krebs CC.
• It can be efficiently operated at a relatively lower feed pressure in comparison to commercial CCs.
• Water recovery in the U/F is low, and so the amount of fine particles entrained into the U/F is small.

In summary, a better size classification was achieved by the JKCC with a low water recovery in the U/F that would benefit downstream processes.