Predicting In-Situ Rock Fracture Parameters Using Soft Kriging

Society for Mining, Metallurgy & Exploration
Y. C. Kim
Organization:
Society for Mining, Metallurgy & Exploration
Pages:
16
File Size:
581 KB
Publication Date:
Jan 1, 1989

Abstract

Three rock fracture parameters define three points along empirical cumulative size (of rock fragments) distribution curves at Inspiration copper mine. These parameters are: rock quality designation (RQD), longest piece of intact rock and % of +1 inch. The RQD and the longest piece are actually measured on drill hole cores whereas % of +1 inch is visually estimated on the same cores. These size distribution curves are then utilized in a variety of engineering applications such as; slope instability analysis, blast design, copper recovery prediction from heap leaching, etc. The three rock fracture parameters for two dominant rock types (porphyry and schist) in the pit had to be estimated separately due to different rock fracture characteristics.' Both ordinary kriging (OK) with measured hard data only and soft kriging (SFTK) with hard and soft data were used to estimate each parameter. The cross validation of 100 measured sample values served as the-basis for evaluating the potential benefits of SFTK over OK. Use of subjective information during SFTK required obtaining (prior) probability distributions for all three rock fracture parameters. These distributions were obtained from the typical rock fracture parameters observed in drill hole cores falling inside the geologically interpreted structural domains with varying degrees. of fracture intensity and also through direct mapping along the exposed bench faces. Two different types of soft information were developed for this study; namely, Type "A" and Type "K". The former is based on sample distribution of fracture parameters whereas the latter is based on two step estimation designed to narrow the range of variation observed in data Type "A". The results of the cross validation were analyzed using a linear regression of actual values versus estimated values. The results confirm an improvement on the estimates made by SFTK with hard and. both types of soft data over those produced by OK with hard data only.
Citation

APA: Y. C. Kim  (1989)  Predicting In-Situ Rock Fracture Parameters Using Soft Kriging

MLA: Y. C. Kim Predicting In-Situ Rock Fracture Parameters Using Soft Kriging. Society for Mining, Metallurgy & Exploration, 1989.

Export
Purchase this Article for $25.00

Create a Guest account to purchase this file
- or -
Log in to your existing Guest account