The recent release of GEOVIA Surpac™ 6.7 features design improvements for grade estimation by enhancing search ellipses and weighting length of composites. Significant increases in the processing speed of the block model enable geologists and engineers to see faster results of larger models with greater coverage or finer resolution. With multi-threading, Surpac 6.7 uses all of the available CPU cores to significantly increase the speed of high resolution modeling, analyzing much larger models in a fraction of the time. Its optimized performance allows the use of much larger models with very high sampling density, which has historically been difficult to collect and analyze in a single model.
Here is a tip that will make better use of composite data when working in Surpac 6.7.
When a composite file is generated from the database within a domain, we always allocate a threshold for the length of sample to be included for the calculation. This results in the creation of composite length below the fixed length set up for compositing. This value is stored in the description field 6 of the composite string file for a single assay field.
For estimation purpose, it is an assumption to consider all composite to be at the same length while in reality they are not. So to affect each composite with its real support, Surpac 6.7 has ability to use the “Weight by field” to select the composite length field. To use it, tick “Weight by field” and select the composite length field.
The result of the estimation is additionally weighted by the value of a DC field of a string. If there are many samples with a small composite length, when this option is selected (and Desc field is set to the D field that contains the composite length), the small samples do not have a disproportionate influence on the estimated result.
By using this method, Surpac 6.7 users will use all the available samples for a better estimate of narrow deposits.
Check out our other posts on Modeling a Fault in Surpac and Why is Surpac so dogmatic about validation of solid models.