ERA

Download the full-sized PDF of Mineral Resource Classification and Drill Hole Optimization Using Novel Geostatistical Algorithms with a Comparison to Traditional TechniquesDownload the full-sized PDF

Analytics

Share

Permanent link (DOI): https://doi.org/10.7939/R3VT1GV9M

Download

Export to: EndNote  |  Zotero  |  Mendeley

Communities

This file is in the following communities:

Graduate Studies and Research, Faculty of

Collections

This file is in the following collections:

Theses and Dissertations

Mineral Resource Classification and Drill Hole Optimization Using Novel Geostatistical Algorithms with a Comparison to Traditional Techniques Open Access

Descriptions

Other title
Subject/Keyword
Drill hole optimization
Drill hole spacing
Cross validation variance
Simulation
Moving window classification
Resource Classification
Neighbourhood restrictions
Geostatistics
Kriging variance
Type of item
Thesis
Degree grantor
University of Alberta
Author or creator
Silva, Diogo S F
Supervisor and department
Boisvert, Jeffery (Department of Civil and Environmental Engineering)
Examining committee member and department
Askari-Nasab, Hooman (Department of Civil and Environmental Engineering)
Wilson, Ward (Department of Civil and Environmental Engineering)
Department
Department of Civil and Environmental Engineering
Specialization
Mining Engineering
Date accepted
2015-01-13T13:54:11Z
Graduation date
2015-06
Degree
Master of Science
Degree level
Master's
Abstract
The classification of mineral resources must follow standards that were created to regulate the public disclosure of projects assisting investors and their advisers in making investment decisions and preventing the publication of erroneous, misleading and fraudulent information. The definition of classification categories are subjective and based on the degree of confidence in geologic continuity, granting the choice of an adequate technique for classification to an expert professional, commonly referred as a competent or qualified person. Many techniques have been developed for resource classification in recent years and to understand the state of practice of resource classification, a survey of Canadian NI 43-101 reports was conducted. The survey revealed that geometric techniques dominate the techniques used for classification and that, although geostatistical techniques are not commonly used in practice, kriging variance appeared as criteria for classification more often than expected. Geostatistical techniques have the potential to introduce relevant information to the classification paradigm, such as, accounting for the spatial correlation of attributes of interest or even allowing the assessment of local distributions that enable the use of meaningful probabilistic classification criteria. Kriging variance is known to generate undesirable artifacts (bullseyes) and often requires post processing. A novel cross-validation variance technique that keeps the advantages of variance based techniques while reducing artifacts is proposed in this thesis. The classification is performed by (1) removing one or more drill holes with highest kriging weight (2) calculating KV using the surrounding data and (3) applying a threshold for classification. The thresholds applied are naturally higher than those originally used for regular kriging variance due to the removal of nearby drill holes. A second technique based on a moving window classification applied to conditionally simulated realizations is also proposed. This addresses the problem of the scale of classification and artifact generation leading to a high resolution classification with reduced artifacts. Moreover, simulation uses meaningful probabilistic criteria for the classification such as precision and confidence (e.g. a block is classified as measured if its grade falls within ±15\% of the mean 95\% of the times). The optimum location of infill drill holes is also addressed in this thesis. An objective function that maximizes classified resources while minimizing the kriging variance is proposed. The optimization algorithm based on an intelligent random search with a random restart and local refinement. Although the proposed technique is not guaranteed to find a global optimum, the proposed methodology is capable of finding reasonable solutions that lead to improved resources. All techniques developed in this thesis are applied to synthetic examples and a case study. The case study is a Cu-Mo deposit located in northern Chile.
Language
English
DOI
doi:10.7939/R3VT1GV9M
Rights
This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for the purpose of private, scholarly or scientific research. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.
Citation for previous publication

File Details

Date Uploaded
Date Modified
2015-06-15T07:03:42.582+00:00
Audit Status
Audits have not yet been run on this file.
Characterization
File format: pdf (PDF/A)
Mime type: application/pdf
File size: 8006877
Last modified: 2016:08:04 02:56:19-06:00
Filename: Silva_Diogo_SF_201501_MSc.pdf
Original checksum: 94b4b33305f68f10bd6da629044a3975
Activity of users you follow
User Activity Date