|
|
Uniform Conditioning - Global G/T Curves Create global grade-tonnage curves for each Selective Mining Unit |
Uniform Conditioning - Global G/T Curves
To access this dialog:
The Global G/T Curves screen, part of the Uniform Conditioning wizard, is used to generate g/t curves for each Selective Mining Unit (SMU). You define the dimensions for mining units on this screen, as well as the cut-offs . This process will use the weighted samples file generated during the Decluster phase to automatically generate a Graph (a new, standalone Plot window sheet) showing the variance between predicted recovered tonnes and grade at different cutoff grades for different SMU sizes.
|
It is not necessary to specify an input block model in order to generate global Grade-tonnage curves. The Uniform Conditioning wizard will allow the Gaussian Anamorphosis and Support Correction options to process different sizes of SMU without the need to specify an input model. |
Generation of Grade-Tonnage information is completed using two distinct processes: first, values are transformed such that the values are gaussian in terms of distribution (the transformation is inversed to get the values back to the original space) - this process is known as Gaussian Anamorphosis and is run once.
The second process is run against each SMU and is known as support correction. The algorithmic background of this process is outside the scope of this Help file, but in summary, support correct is where a multivariate change of support (point to SMUs) is performed for the histograms.
In the geostatistical sense, recoverable reserves means the ore tons and metal quantity contained in those tons for a given SMU. The average recoverable grade is calculated as the metal divided by the tonnage.
Instead of dealing in absolute tons, uniform conditioning considers the tonnage as the number of SMU's whose grade is above the cut-off grade . The metal is calculated as this tonnage multiplied by the SMU grade. This type of formalism is used to express the two variables in terms of indicator variables.
To calculate the absolute tonnage and metal, these indicator variables are multiplied by the SMU volume and the density of the ore. However, the potential problem here is that an assumption is made that, at the time of mining, the real SMU grade is known. This is not true as this information is only revealed during production where an estimate of this grade is known and calculated from production samples.
The SMU's sent to the mill as ore are those whose estimated grade, not necessarily real grade, is above the economic cut-off. So some non-waste blocks will be incorrectly classified as waste and some waste misclassified as ore. The amount of misclassification is due to the so-called information effect. For a non-conditionally biased estimated the two types of misclassification will at least annihilate each other in terms of tonnage but not in terms of recovered grade.
Find out more about the Information
Effect...
Field Details:
The Global G/T Curves screen contains the following fields:
No. of discretisation
points: discretisation allows you to simulate a three dimensional
array of points, distributed regularly within the SMU. As Uniform
Conditioning adopts a kriged
approach to estimation, the discretised points are used for calculating
the covariance of the cell
with each of the surrounding samples. This is then used in calculating
the kriging weights. You can use the spin buttons in this area of
the screen to set the number of discretisation points in X, Y and/or
Z for each SMU.
If an even number of points in a direction are defined, then the points
will be spaced around the centre line. If an odd number of points
are defined then there will be a point on the centre line and the
others will be spaced regularly towards the edges. Find
out more about discretisation...
Block orientation angle: in the case where one or more rotations are required to describe block anisotropy, you can enter X, Y and/or Z values here. For example:
X=20
Y=60
Z=40
This relates to first rotation is a dip of 40 degrees around the new
X-axis, 60 around the new
Y-axis and a conventional azimuth rotation of 20o around the Z-axis.
Normalise variogram sill (and information effect variance and covariance): due to the mathematical complexities of the kriging calculations it can sometimes happen that the kriged variance is slightly greater than the sill of the model variogram. Select this option if you intend to use a normalized variogram sill values during grade-tonnage curve generation. As described, this will also normalize the variance and covariance calculations from the information effect.#
If this check box is enabled, the variogram will be normalised and, if disabled, the input variogram won't be normalised by the uniform conditioning process.
Information effect: with Studio's Uniform Conditioning module, you can (optionally) incorporate the information effect to the estimation of the grade tonnage curves: during the production stage, the actual grades are recovered and may then be taken into account so the decision between ore and waste is made upon more accurate estimates of the SMUs. Therefore you can anticipate future decisions before obtaining the production blast-holes results, because only the kriging variance of these SMUs final estimates is necessary.
Use one of the following options to determine how (or if) the information effect described above will influence the grade-tonnage curve display using one of the following options:
-
Ignore: select this to ensure the information effect will not influence the g/t information that is generated when curves are created.
-
Manually enter: if you are aware of a variance/covariance value you can select this option and enter them here.
-
Calculate: choose this option to ensure the information effect is considered. You will need to enter the planned sample spacing and number of samples in XYZ directions. A variance-covariance will then be calculated automatically. Note that if you are entering these values manually, you must ensure that the sample spacing (which represents the panel dimensions) must be greater than or equal to the largest SMU block size - it is not possible to perform Uniform Conditioning where any SMU dimension exceeds the dimensions of the panel with which it is associated.
Find out more about the Information Effect...
Create global G/T curves: calculate Grade-Tonnage curves at the specified SMU dimensions and parameters.
View global G/T curves: show the generated G/T curves in the Studio Plots window on an automatically created chart Sheet. Note that you can use all of Studio's charting functionality to further interrogate/format the generated chart. You can find out more about charting in Studio using the Help menu on the left (Using Studio 3 | Presentation & Reporting | Charts)
Copyright © Datamine Corporate Limited
MIN 20044_00_EN
References:
Assibey-Bonsu W and Krige DG, Use of Direct and Indirect Distributions of Selective Mining Units for Estimation of Recoverable Resource/Reserves for New Mining Projects, In Proceedings from APCOM'99 Symposium, Colorado School of Mines, Golden, Colorado, pp 239-247, 1999.
Lantuéjoul C, Cours de séléctivité, Course notes C-140, Centre de géostatistique, Ecole des Mines de Paris, Fontainebleau, France, 72 pp, 1990.
Matheron G, Forecasting block grade distributions: the transfer functions, In Advanced Geostatistics in the Mining Industry, M Guarascio, M David & C Huijbregts, eds, Riedel Publishers, Dordrecht, pp 237 - 251, 1976.
Matheron G, The selectivity of the distributions and the "second principle of geostatistics", In Geostatistics for Natural Resource Characterization, G Verly, M David, AG Journel & A Maréchal, eds, Riedel Publishers, Dordrecht, pp 421 - 433, 1984.
Remacre AZ, Uniform conditioning versus indicator kriging: a comparative study with the actual data, In Geostatistics: Proceedings from the third international Geostatistics Congress, Sept 5-9, 1988, Avignon (France), M Armstrong, ed., Kluwer Academic Publishers, Dordrecht, pp 947 - 960, 1989
