Letters: COMPUTER BUGS

Computers are said to have the following drawbacks:

* Computer-based reserve models were developed for porphyry copper deposits and hence are not applicable to gold deposits, especially lode vein deposits mined underground.

* Computer-plotted maps lack the human touch of manual plotting.

* Computers ignore the geologist’s knowledge base, and ore reserve estimation becomes strictly a number-crunching process.

* The computer leads the ore reserve calculation and prevents the geologist and mining engineer from working as a team.

These concerns stem largely from the still-prevailing opinion that the computer is an omnipotent black box providing magic answers. Nothing could be further from reality. The computer is a powerful tool but only when properly applied. Its application does not release mining practitioners from applying the same principles that they otherwise would apply to a manual ore reserve estimation.

In the early 1970s, most computer programs for reserve computation contained mistakes or “bugs.” However, after almost 20 years of testing, program bugs or errors that would have any effect on the actual reserve determination are by far the exception rather than the rule, and the “bug” excuse is rarely an applicable criticism.

Computer-aided drafting (CAD) systems, such as Autocad, allow practitioners to fully utilize their geological data (i.e., the human touch) while enjoying the computer’s processing capabilities. With CAD, a conventional manual ore reserve can now be completely automated. The principles are identical; the difference is that, instead of using a planimeter, the computer calculates areas. CAD-based and manual ore reserves rely on data presented in two dimensions (cross-sections and level plans). Enhanced precision in the determination of the geometry of an ore deposit is obtained by the application of a true 3-dimensional approach, such as the 3D Solids Component Modelling technology developed by Lynx Geosystems.

Computers significantly reduce the time and effort spent in updating reserves. Far from preventing geologists, mining engineers, geostatisticians and computer specialists from working as a team, the computer can improve co-ordination among these professionals. If teamwork does not exist, the fault probably lies with the practitioners, not the computer.

The question of appropriateness of computer-based reserve models for underground gold deposits is an important one. Again, teamwork is needed. To provide a definitive answer to the question above, the following studies are required:

* comparisons between the actual production results and the reserve estimates made before production.

* detailed accounts by geologists and engineers of their experience with computerized ore reserves in underground operations.

CAD allows the imposition of tight geological constraints on geostatistical models to ensure the correlation between grade values and geology is respected. Overestimation of the grade of gold deposits with high nugget effects is frequent, causing disastrous production results. It is therefore important that the mining industry continue to improve grade estimation techniques. The combined geological and geostatistical models have the potential to resolve this.

How to reduce computer misuse? First and most importantly, the level of education among geologists and engineers must be raised. Second, the experience gained by users needs to be documented.

And last, three important principles:

* Professional judgment should always prevail over computer output, since the latter is simply a complement to the practitioner’s judgment, not a replacement.

* Program standards should be used to ensure that the same result will be produced given identical geological and mining data. Currently, there are no program standards in the exploration and mining industry, and it should be the role of professional associations to help set these standards. Commonly used programs include Autocad and GEOEAS, a public domain geostatistical package.

* Procedures need to be established, documented and taught for professionals within an organization who are calculating ore reserves for the first time. This is especially true if reserves are computed in offices with individual microcomputer systems.

From the perspective of a mining decision-maker, important questions are: Do these sophisticated computer-based techniques offer a better standard than the traditional manual approaches? Do they provide more accurate estimates? Do they improve the efficiency of mining operations?

To help answer these questions and to provide an opportunity for users to present examples of the problems encountered and the solutions adopted, the sponsoring societies of CAME (Computer Applications in Mineral Exploration) are holding a workshop on “Computer Treatment of Exploration and Mining Data — Dos and Don’ts” on March 9 and 10, 1990, as a pre-convention activity to next year’s Toronto convention of the Prospectors and Developers Association of Canada.

William Pearson, Derry, Michener, Booth & Wahl Normand Champigny,

The Coopers & Lybrand Consulting Group

]]>