|
|
| |
Quantifying
risk to reduce and manage uncertainty in rehabilitation sign-off for closure
Inaccurate predictions of the size or grade of an orebody made from
limited drill core data can have major operational and economic implications
for a minesite. Modern geostatistical modelling procedures can provide
improved predictions of the spatial variation in key attributes and a
quantification of the uncertainties involved in these predictions.
Like an orebody, minesite rehabilitation is also spatially variable and
extrapolations of rehabilitation success from a small number of monitoring
points or transects are similarly challenged by uncertainty. Here, the
uncertainty may manifest as a reluctance on the part of the government
regulator to sign-off on rehabilitated areas at mine closure, because of
unknown ‘risk’.
This collaborative project with the W.H. Bryan Mining Geology Research
Centre at UQ is exploring the application of cutting-edge geostatistical
modelling tools to better predict the variation in key attributes in areas
to be rehabilitated and to quantify the likelihood of long-term
rehabilitation success. It is hoped that by quantifying the probabilities of
success, or the likely risk of rehabilitation ‘failure’, more informed and
confident decisions about rehabilitation sign-off can be made.
Studies of the approach on test datasets from mines in Queensland and New
South Wales have so far proved promising. The modelling techniques used in
orebody definition appear robust, but further work is needed in
strengthening the links with vegetation thresholds for multiple attributes,
and to take account of changes in spoil and soil properties over time in
these newly developing systems.
Sponsors: Minerals Industry and Queensland State Government
through the SMI |
|
