By Antonio Garibaldi,  Zak Hynd and Ignacio Torresi.

This is written for those who are having trouble modelling veins with fragmented data, especially channel samples.

Leapfrog Geo’s vein tool was designed to work with drillhole data, ideally drillholes which intersect both walls of the tabular vein structure. If you are using the vein modelling tool with fragmented channel sample data you will find these tips and tricks helpful.

How the vein tool works

To begin with, it is worth understanding how the vein tool works. See figure 1 below.

Figure 1. The five basic construction steps for a vein.

Figure 1. The five basic construction steps for a vein.

1. First, a reference surface is generated from the midpoints of the vein intervals.
2. Vein segments are assigned Hangingwall (HW) and Footwall (FW) sides based on their orientation to the reference surface.
3. Separate HW and FW points are generated at the ends of these segments.
4. The HW and FW surfaces are generated – both are offsets of the reference surface which fit to the respective points.
5. The final product is the volume enclosed between the HW and FW surfaces.

Vein segment classification

Due to the way Leapfrog Geo automatically classifies vein segments, we can define three basic types of vein samples.
• Wall-to-wall samples are in the same hole or channel and represent/touch/intersect both walls of the true vein structure.
• Incomplete vein samples are samples which represent only one wall of the true vein structure.
• Internal vein samples do not intersect any of the true vein walls.

Figure 2. Groups: Right = wall to wall samples, Left = incomplete samples, Centre = internal samples.

Figure 2. Groups: Right = wall to wall samples, Left = incomplete samples, Centre = internal samples.

The ‘true’ vein structure is represented by the light-green colour, with the walls represented by the brown lines. Green intervals represent logged vein samples, purple intervals represent logged ‘outside vein’ samples, the grey lines represent the unsampled sections of the drillhole or channel.

As you can see from figures 3 and 4, Leapfrog Geo’s automatic vein segment classification does a good job with wall-to-wall sample types – even when they are highly fragmented like the last hole – but doesn’t do so well with incomplete and internal sample types. These incomplete and internal sample types can produce poor surface triangulations because the surfaces are contacting each of their respective segment end points. The hangingwall and footwall surfaces can intersect one another, generating holes in the modeled volume of the vein.

Figure 3. Results of Leapfrog Geo’s automatic vein segment classification. Each segment end is represented by one of three types: Hangingwall (red), Footwall (blue), and Excluded (grey). The resulting vein wall surface triangulations are green in this image.

Figure 3. Results of Leapfrog Geo’s automatic vein segment classification. Each segment end is represented by one of three types: Hangingwall (red), Footwall (blue), and Excluded (grey). The resulting vein wall surface triangulations are green in this image.

Figure 4. The ideal classification of these vein segments. Internal samples are all classified as ‘Excluded’. The internal ends of the incomplete segments are excluded.

Figure 4. The ideal classification of these vein segments. Internal samples are all classified as ‘Excluded’. The internal ends of the incomplete segments are excluded.

Manual editing

With some manual edits, you can achieve the ideal vein segment classification in Leapfrog Geo.

Edit vein segments

To correct the vein segments on the incomplete samples you will need to manually edit the vein segments and override their automatic classifications. Right-click on the vein segments and select Edit In Scene (see figure 5).

Figure 5. To edit vein segments, right-click on the Vein segments object underneath the vein in the project tree.

Figure 5. To edit vein segments, right-click on the Vein segments object underneath the vein in the project tree.

Click on a vein segment, then in the Vein Segment Orientations dialog uncheck Auto for the point (A or B ) which is incorrect. Set the point to the correct classification. In the example illustrated in figure 6, the partial segment’s point A has been excluded so it will be ignored by the vein wall surfaces. Repeat this for all the incorrectly classified vein segments from partial samples.

Figure 6. Manually override the vein segments classification.

Figure 6. Manually override the vein segments classification.

Ignore internal samples

It can take a long time to manually override the vein segment classification, especially with internal vein samples which need to have both segment ends (Point A and Point B) excluded. One way to avoid this is to ignore these samples with a query filter. First, you will need to classify the internal samples in the interval table. Then create a new Interval Selection on the interval table which your vein is built from (see figure 7).

Figure 7. Create an interval selection on the vein interval table.

Figure 7. Create an interval selection on the vein interval table.

Select and assign all the internal samples to a new ‘lithology’ code (see figure 8). See help documentation for more information on the interval selection tool.

Figure 8. Using the Interval Selection tool, internal vein samples are selected and classified.

Figure 8. Using the Interval Selection tool, internal vein samples are selected and classified.

Once the interval selection column is created, on the same interval table create a new Query Filter which ignores the internal samples.

Figure 9. Create a new query filter which ignores the internal vein samples.

Figure 9. Create a new query filter which ignores the internal vein samples.

To apply this query filter to the vein, open the vein segments, uncheck the option to inherit the query filter from the GM (geological model), and select the new query filter from the drop-down list.

Figure 10. Apply a query filter to the vein segments to ignore the internal samples.

Figure 11. Leapfrog Geo’s vein segment classification results after ignoring the internal samples.

Figure 11. Leapfrog Geo’s vein segment classification results after ignoring the internal samples.

Ignore points at the end of holes

Depending on the data, it may be possible to reduce the number of manual edits by automatically excluding the segment ends at the end of a hole or channel. The default setting for vein segments is to include points at the end of holes. To change this setting, open the vein segments by double-clicking (or right-click and select Open), highlighted in orange in the image, then uncheck the option to include points at the ends of holes (see figure 12).

Figure 12. Uncheck the option to include points at the ends of holes.

Figure 12. Uncheck the option to include points at the ends of holes.

Figure 13. The automatic vein segment classification which excludes points at the ends of holes.

Figure 13. The automatic vein segment classification which excludes points at the ends of holes.

As you can see from figure 13, vein segment ends are excluded at the end of the holes or channels. This has solved the classification problem for some of the incomplete and internal samples. However, if there are a lot of vein samples which extend to the end of holes (e.g. hole 11), or if the vein represents a grade shell which should tightly enclose all vein samples, then changing this setting may not always be appropriate.

Conclusion

With respect to Leapfrog Geo’s automatic vein segment classification, the most recommended vein sample type to use is wall-to-wall. Even if the wall-to wall sampling is fragmented (separated by unlogged intervals or non-vein intervals), the automatic segment classification will produce appropriate results. With channel samples, if the vein samples can be included in the same continuous channel, perpendicular to the tabular vein structure, you can minimise or eliminate the need for manual edits in Leapfrog Geo (figure 14).

Figure 14. Fragmented channel samples. The samples on the left will need to be manually edited in order to produce a reasonable vein triangulation, whereas the samples on the right will work automatically.

Figure 14. Fragmented channel samples. The samples on the left will need to be manually edited in order to produce a reasonable vein triangulation, whereas the samples on the right will work automatically.

Future vein modelling improvements

We intend to continually improve the user experience of Leapfrog Geo based on user feedback. Comment below and let us know what you think.