In this installment, some aspects related to the spatial integrity of the sample will be discussed. Carrasco, Chilès, and Séguret (2008) highlight this aspect: “Sometimes, in order to get enough weight to perform several tests, several increments coming from very different locations are combined in a composite sample, so that the spatial location is lost.”
First of all, let’s see where the problem comes from. When working with core samples, typically 50% is allocated for geology, 25% remains for future reference, and 25% is intended for metallurgical testing (Figure). Furthermore, using 15 meters of core in HQ diameter (63.5 mm), it is possible to obtain approx. 25 kg of material for metallurgy testing.
On the other hand, geometallurgical testing may include SMC Test® (5 kg), Bond Work Index (7 kg), grinding kinetics (5 kg), rougher flotation (1 kg), lithogeochemistry and mineralogy (1 kg). The mass required for this characterization is approx. 19 kg (without considering QAQC samples and/or repetitions).
Previous aspects suggest that the mass available for metallurgical testing may be insufficient without careful planning. One strategy to deal with this problem is simply taking additional core increments from locations with similar geologic attributes (Figure). In the exemplified case, having 25 m core length results in approx. 40 kg for metallurgical tests.
A disadvantage of the previous approach is that the composited sample provides an average result, destroying the spatial variability and the possibility of differentiating the performance of individual segments. This point is relevant since there is variability in the metallurgical response even within the same type of rock.
In practice, if the sample is a continuous segment of core (e.g., 15 m @ ¼ HQ) it is called a punctual sample or variability sample. In the case the sample is formed using intervals without spatial continuity, it is called a composited sample. In this way, it is easy to address whether the spatial integrity of the sample is preserved (punctual sample) or has been compromised (composite sample).
In the next installment, we will see other practices that compromise the spatial integrity of the sample, limiting the capture of metallurgical variability.
What is your experience with this or other types of practical problems when assessing metallurgical variability? Let’s talk, write to contacto@m2p.cl.
References of this publication:
Armstrong, Margaret (1998) Basic Linear Geostatistics. Springer. Berlin, New York. 155pp.
Goovaerts, Pierre (2016). Sample Support. 10.1002/9781118445112.stat07754.pub2.
Ljung, Lennart (1998). System Identification, Theory for the user (2nd Edition).
Pedro Carrasco, Jean-Paul Chilès, Serge Antoine Séguret. ADDITIVITY, METALLURGICAL RECOVERY, AND GRADE. 8th international Geostatistics Congress, Dec 2008, Santiago, Chile. pp.on CD. hal-00776943
