Company Announcement - Although there has always been the need to obtain representative samples in mineral particulate systems, until recently there has not been a proven technology that can take a representative sample from pipelines for horizontally pressurised tailings streams. This is according to Rolf Steinhaus, sampling specialist at Multotec Process Equipment, who says that incorrect sampling solutions are often employed on the tailings section of mineral process plants and particularly so in gold, platinum and base metal industries.

“Metallurgical accounting is very difficult when the data or results produced are from these inaccurate arbitrary material collections. Meaningful metallurgical accounting is essential, as plant operators need to have accurate information as to what comes into and leaves a plant in order to establish how efficient the actual beneficiation process is and what yields or recovery can be expected of the metal or precious mineral. Reliable reporting and corporate governance calls for representative sampling, using correctly designed sampling equipment and implemented protocols that will produce accurate data,” Steinhaus points out.

Multotec has developed a mechanical design for the True Pipe® sampler and, together with Dr Paul Roberts, has taken out a patent on the product. The prototype has been extensively laboratory tested and the results compared favourably with a correctly designed vezin sampler used as a control. Data generated has been very encouraging and Multotec is now in a position to consider industrialising the unit together with a partner.

This sampling system comprises pipework and valves that will isolate and correctly delineate a sample increment under controlled conditions. Operation of the valves is synchronised, resulting in symmetry between two separate biasing mechanisms occurring in the vicinity of the two relevant valves which will cancel each other out. What is misplaced into the sample at one end is displaced from the sample at the other. “The principle of symmetry was statistically proven for this particular design. The True Pipe® sampler will ultimately have application in high pressure slurry lines with pressures of up 2 000 kPa and slurry line velocities of up to 6 m/s, however testing for less challenging applications needs to be done,” says Steinhaus.

During the development stage and as a separate investigation, the stream was carefully split into two 50% streams. “Statistically the results re confirmed that due to segregation effects, partial stream sampling is not advisable  and that splitting the main stream into partial streams is once again (like any sample stream diversion ) never an acceptable shortcut when sampling representatively. Prior to the development of this innovative sampling system for tailings pipelines, all measurements taken would be considered as non-representative and would therefore have been biased,” he adds.

Plant tailings streams are generally pumped out in pipelines to slurry and tailings dams, a process which typically does not lend itself to conventional cross cut sampling methods. Current sampling devices include injector or poppet samplers, which are pneumatically driven arms that collect small increments of fluid from the centre of the stream via an opening in the shaft. Sampling practices of sampling part of the stream, part of the time is not acceptable or at all reliable for heterogeneous particulate systems.

“One needs to be able to take a full cross cut or slice of the stream, due to the fact that segregation of material is always present in particulate systems. This methodology provides an accurate and precise composite sample if cuts are taken frequently,” says Steinhaus.

Another unsatisfactory sampling system being used is the T-piece bleed off valve which bleeds off a portion of the stream. “This is an ad hoc material extraction via a valve often situated arbitrarily on horizontal pipe columns and results will differ depending on the location of the device around diameter of pipe. Larger heavier particles will be collected preferentially if it is situated at the bottom of the pipe. Again, this is not representative of the whole but regrettably, in the absence of anything better, users often believe that the imperfect sample result will suffice,” Steinhaus says. Numerous sampling errors occur, with resultant and potentially poor control decisions with this method of specimen collection due to the wholly inadequate and limiting design of this device.

The pressure pipe sampler is another example of a typically used slurry collection device. Part of the stream is sampled continuously via a probe within the dilated housing of the sampler body pipeline. Steinhaus points out that it is incorrectly assumed that the turbulence which occurs in the dilated section of the pipe with or without turbulence bars, will overcome sample delimitation and extraction errors due to some homogenisation. This is definitely not the case as it is still not a full cross stream sample.

Endemic to these three slurry collection devices is their inability to provide a representative and unbiased sample. “However, because they are all low-cost devices, they are often considered an attractive option. In simple terms, the widespread use of these devices underpins the lack of understanding of sampling theory, with the principles of cross stream sampling which have been proven,” adds Steinhaus.

“One needs to remember that sampling of a stream representatively particularly in a pressurised environment is a complex process. There are up to six or more types of sampling errors that can occur when trying to extract a sample increment from a whole stream or a portion of the particulate stream. The True Pipe® sampler addresses the majority of these challenges and provides a sample of integrity under controlled conditions,” Steinhaus concludes.