Quality Appraisal for Geophysical Borehole Logs

The aim of this project was to investigate efficient means for identifying suspect borehole logging data from hundreds or even thousands of boreholes. The intent was to classify data as acceptable, questionable, or unacceptable. The envisaged ultimate benefits of the project were

• fewer misleading interpretations based on erroneous data,
• improved mine models incorporating quantitative information from wireline logs, and hence
• enhanced productivity and reduced risk.

Main Achievements and Deliverables

The quality appraisal (QA) methodology developed in the course of this project distinguishes three types of criteria:

• generic criteria, which can be applied to single logging runs in individual holes, and hence to any logs;
• repeat hole criteria, which apply to multiple logging runs of a single hole; and
• production hole criteria, which apply to single logging runs in multiple holes.

Log data quality is assessed via application of an appropriate sequence of statistical criteria. The approach guarantees consistency of log data, but not its absolute accuracy. Each mine site can prescribe the level of consistency which is acceptable. In order to test the practicality of the approach, proto-type QA software has been developed. The initial release version of the software is ripe for refinement and further development. The software is modular in structure, to facilitate incorporation of additional utilities in the future. Site-specific criteria could be developed, if required.

The QA methodology and the proto-type program developed in this project can be applied to historical data, or to new data “at the truck”.

This report describes the methodology, and illustrates its application via the proto-type software.

The following recommendations have been formulated:

• Trial the proto-type LogQA software at a number of mine sites. The trials will highlight any general limitations of the software, and provide an opportunity to define options for integration of QA into routine data processing.
• Authors will endeavour to make themselves available, on a commercial basis, to expedite the QA software trials.
• Examine QA of new data as well as historic data during the on-site trials.
• Retain LAS as the ‘standard’ data format for QA.
• Define and adopt a standard LAS file header
• Establish calibration holes, to underpin data accuracy.