SWATH PLOT VALIDATION with GEMS

The importance of the Validation of the Block Model Estimation with GEOVIA GEMS: SWATH PLOT VALIDATION

Please note: This post is also available in Spanish – Click Here

 

When estimating a block model, no matter what interpolation method is chosen, it is essential to validate the obtained results in order to ensure the estimation process quality according to the current industry standards.

By validating the block model, we look for:

  • That the estimation be globally unbiased
  • Minimize local bias
  • Keep the estimation smoothing at a reasonably acceptable level

 

There are numerous methods to do this and through upcoming blogs we will discuss these three (3):

  1. Graphical validation.
  2. Statistical validation.
  3. Swath Plot validation.

Finally, we will explain the method of Swath Plot Validation:

 

SWATH PLOT VALIDATION OF THE BLOCK MODEL ESTIMATION

Objective:

Validate that the local bias of the estimation is acceptable, i.e., that the local means of the estimated blocks are similar and follow the same trend as the “real” data (declustered composites) for the 3 main directions of the space.

How to do it?
Slices are defined in the three main directions of the space. Within each slice, the mean of blocks and of declustered composites are calculated. Subsequently, those calculated mean are plotted in the three main directions of the space.

 

Care to take into account at the execution:

  • Perform the analysis in the three perpendicular directions of interest, such as main directions of anisotropy
  • That the composites have the same support as the block size
  • Carry out the analysis only on the best estimated blocks, for example, those candidates for measured and indicated resources
  • That the composites have been previously declustered. With that purpose, it is suggested to apply Nearest Neighbor (NN) declustering method
  • That the analysis be performed separately for each estimation domain

 

Expected results:

  • That the curve of the mean of the estimated blocks follows the same trend as the curve of the mean of the declustered composites.
  • That the curve of the mean of the estimated blocks be smoother than the curve of the mean of the declustered composites, but that it respects the same shape and trend without being exaggeratedly flat.
  • That the curves of the mean of the estimated blocks and of the declustered composites cross each other, preventing the curve of the estimated blocks from passing below or above the curve of declustered composites If this happens, it is a clear sign of under or overestimation.

 

Swath plots are obtained, as follows:

 

  • Know the area on which the slices that will constitute swath plots will be established.

 

 

 

 

  • Divide the work area into slices for the 3 main directions of the space, for example, 100 m wide slices in NS and EW directions, and 80 m wide slices in vertical direction (elevation). Generate closed polylines that represent the slices and within which the basic statistics of the estimated blocks and of the declustered composites (NN block model) will be calculated.

 

Below, it is included an example in the NS direction.

Generation of closed polylines:

 

 

 

 

 

 

 

  • Add an appropriate TAG, which allows to identify the polyline

 

 

 

  • Copy the polyline every 100 m (or every the slice width that has been chosen) in the N direction, until covering the entire area to be analyzed (5,020 m in this example)

 

 

 

 

  • Change TAG field to each of the copied polylines. Select (manually) with mouse the line whose TAG will be changed and then execute the change. Repeat the same process for each of the closed polylines.

 

 

 

 

 

  • Choose the blocks that are inside each of the slices

 

 

 

 

• Report the basic statistics within the previous selection for the estimated blocks and declustered composites (NN)

 

 

 

 

Finally, examples of the resulting swath plots for three directions are shown below:

 

 

 

 

 

Maria Angelica Gonzalez

María-Angélica is a Mining Engineer with a Diploma in Geostatistics and she is a candidate for a Master of Science mention Geology. She is registered as a Competent Person with the Qualification Commission of Competencies in Resources and Mining Reserves of Chile. She started her professional career in 1999 and has worked in consulting and project operations for diverse types of mineral deposits, performing the geostatistical evaluation of elements of economic interest as well as geometallurgical and contaminant variables. She is experienced in metallic and non-metallic projects, such as copper, gold, silver, molybdenum, iron, zinc, lead, titanium sands, phosphates, carbonates, iodine and nitrates, both in America and in Africa. Her current position is Senior Services Manager of Dassault Systèmes Chile (GEOVIA), serving the Mining Industry in Latin America. She is in charge of both consulting and training as well as technical sales, mainly in the technical and administrative leadership of projects. Previously, she worked as Leader of Geology and Resources of the Consulting Group at AMEC. AMEC International Ingeniería y Construcción.

Latest posts by Maria Angelica Gonzalez (see all)