As I write this article I am coming towards the end of 30 years of resource estimation and reporting with SRK. When I joined the available software was 2D based and the interpolation process took a long, long time and was mainly done overnight so as not to tie up a machine all day. Many estimates were still being done using graph paper, colouring pencils and transparency. Hard copy sections and plans were the basis of almost all geological models. While most exploration programmes and new mines were computerised, most mines still operated a paper system and while geostatistics had been “invented” most estimates were still being done using classical methods. 2D inverse distance interpolation and polygonal analyses were the norm and no commercial software system was up to do anything more than simple statistical and geostatistical analysis. For my PhD a few years earlier I wrote the semivariogram and kriging algorithms into Lotus 123 and used a spreadsheet to produce 2D semivariograms and block models. Try and tell that to the young geologists of today, they won't believe you!
Not only was computerised resource estimation in its infancy, resource reporting was poorly regulated 30 years ago. The first JORC Code had been published a couple of years earlier and was starting to gain traction but there were big differences between the various codes in use. Some used the same terms as others but where they did the definitions were different. Fortunately this did not matter too much because most readers did not understand the codes anyway and most reports did not state what code was being used.
Since 1991, the software capability has improved, more elaborate geostatistical
approaches are now open to everyone and modelled orebodies now tend to
look more like mineral deposits and less like Klingon warships. Mineral resource
reporting is much more standardised and continually being refined and improved.
It still remains critical that the estimators understand the orebody geology and the theory behind, and limitations of, the methods they are using, and these are strong themes throughout this newsletter, but notwithstanding this it is clear that we are in a better position now to produce better estimates and to convey the confidence we have in these compared to 1990. I can only wonder what changes the next 30 years will bring.
