The importance of ensuring that the correct slurry sampling design is used for the gold sector cannot be overstated, according to Multotec Product Specialist for sampling Rolf Steinhaus. He, with his knowledge and experience gained in 29 years in the mining sampling sector and 19 years with Multotec, proficiently assists customers with the optimum design for reliable gold metal accounting. The picture shows a cross section of a Multotec launder and vezin layout. A number of aggravating factors come into play when sampling for gold in mineral bearing slurries.
Firstly, the gold is found in such small quantities (ppm) in slurry samples, and has a very high specific gravity of 19.3g/cm3. In addition, gold particles are extremely fine, with measurements of typically less than 20 μm in many instances. Gold will easily segregate and find any opening or crack available to settle and elude detection. Sampling therefore is very specialised and that it needs to be extremely well controlled to derive the benefits of correct equipment designs. Based on this high level of design accuracy, it is fair to say that the reliable design of samplers for the gold sector could effectively be introduced into any other sector.
In the case of slurry samplers, one needs to carefully consider the design of the cutter box in relation to the length and width of the cutter blade, the sample volume it collects for each application as well as the overall geometry of the cutter and how it interacts with the one dimensional process stream when taking a full cross cut primary sample increment. The inclusion of wash water systems and retention hoppers to retain the primary sample for subsequent multiple sub cuts, taken by the secondary vezin type cutter, are recommended features to enhance sample integrity further. Composite sample dewatering in a pressure filter can also considered to reduce sample handling.
Steinhaus believes that in order to achieve proper metallurgical accounting balance within gold plants one needs to undertake cross stream sampling on the leach feed and CIP tailings or total plant tailings. The inclusion of a secondary rotary cross cut sampler, with dual or replicate sample capability, will ensure that samples have the necessary levels of integrity and that sampler performance can be ascertained by monitoring A and B sample precision levels.
Steinhaus cites Multotec’s Two-in-One Sampler® and the Wet Linear Cross Sampling System as two examples where correct design methodologies are able to provide accurate sampling data, provided correct sampling protocols are adhered to by plant staff. A well-designed sampling system, with proper integration into plant regular operator inspections, will provide a high level of accounting integrity, which ensures that unaccounted losses or gains are minimised in the gold plant.
It is always advisable to base sampling design compliance on the recommendations provided by Francis Pitard and other recognised specialists in the mineral sampling field. This input should be further complemented by a close working relationship with industry acknowledged sampling consultants, or in-house sampling champions, conversant with sources of sampling errors for maximised compliance and QA/QC.
Steinhaus points out that there is no definitive textbook on correct sampler design. Published guidelines in certain standards are often sketchy and applications experience must not be discounted.
Every application in the gold industry has unique considerations in terms of material characteristics with regard to density, flow rates, turbulence, slurry line velocities and particle size distribution. This observation is especially relevant to gravity gold circuit sampling, where segregation errors remain a particular challenge due to coarse gold particles. So while there will be certain commonalities in the sampling design procedures, these exceptional characteristics have to be considered.
When sampling any process stream, there is an inherent variability association between portions of that stream. These variations cannot be changed and therefore have to be accepted once reasonably minimised. This variability leads to uncertainty or lack of confidence in the final sample result. However, more critically, if the sampling equipment is incorrectly designed, one will face significant errors or biases which is generally far more significant.
The two biggest errors which could occur are sample delineation/delimitation and sample extraction, as per Pierre Gy’s unified theory of sampling, and are directly related to sampler design and stream interaction. More than any other errors, these two factors contribute most to poor sampling results and the generation of inaccurate or misleading data.
Ultimately, any gold industry sampling system that is not based on cross stream sampling methods, where every particle has an equal opportunity of being sampled in the flowing process stream, is destined to provide inaccurate sampling data and lead to poor decision making. Aligning oneself with a reputable supplier, who can leverage years of experience in the gold sampling design sector, will afford management every opportunity of an accurate metallurgical accounting balance.
