SPATIAL VARIABILITY IN A WASTE ROCK PILEA new feature in SVFLUX Professional (part of SVOffice 2009)February 20, 2009
In general, quantifying sources of uncertainty has always been challenging. The list of potential aspects of introducing uncertainty into a numerical model includes: laboratory uncertainty, uncertainty regarding the governing process in a numerical model, climatic uncertainty, and uncertainty related to the proper discretization of a numerical model. Spatial variability is of particular concern in performing seepage modeling because there is often uncertainty regarding the ability to measure the hydraulic conductivity (both saturated and unsaturated) of the material at the site. It is possible that laboratory measurements of saturated hydraulic conductivity could be in error by up to two orders of magnitude. If samples are taken around a site, even in the same layer, the amount of variance in saturated hydraulic conductivity can be significant. The current methods for dealing with this type of uncertainty are often associated with Monte Carlo type techniques, which will vary the material properties through a normal distribution and perform hundreds or thousands of numerical model runs. The draw-back for this method is that, in each trial run, the material properties for the target region are assumed to be homogenous.
This new feature is implemented in the latest version of SVOffice 2009 in the SVFLUX module. This is a free upgrade for all existing SVOffice 2009 professional license holders. An example of this application can be seen in the example models of waste rock material on this page. In this particular example, the flow out the toe of the waste rock facility has been noted and measured. The user then therefore desires to set up a numerical model with spatial variation and note the change in the numerical model when including spatial variation. The impact can be noted from the results of this model. |