204
estimation is the kriging variance (KV). The KV takes into account geostatistical parameters which combine both geometric and geological inputs in characterizing the uncertainty associated with the estimated parameter. The consideration of the spatial structure
of the estimated variable and the redundancy between samples are the purposes of using
kriging variance as the criterion for classification. In the process of assigning confidence
levels, estimated blocks with high kriging variance have lower confidence than those with
lower kriging variance. Emery et al (2006) suggest the establishment of KV thresholds for
measured, indicated and inferred mineral resources classification categories based on a
given desired number of drill holes, spacing and the variogram model in a given domain.
Classifying each block into a specific measured, indicated and inferred mineral resource
category is done on the basis of KVs associated with each block and these previously
determined thresholds.
2.2.2 Classification based on 90% confidence interval
This technique involves calculating the 90% confidence intervals (CI) on the tonnage, grade
and metal content within quarterly and annual production blocks either through KV or
geostatistical conditional simulations. A recognized mining industry practice in the application of this technique is the ability to determine the drill hole spacing to be enough in
the prediction of tonnage, grade and metal content within ±15% relative precision at 90%
confidence interval within a quarterly or annual production rates. Measured resource considers quarterly production period and indicated resource corresponds to annual production
period. For the detail description of mineral resource classification based on 90% confidence
intervals, please refer to Verly et al (2014).
3 SURVEY RESULTS AND DISCUSSIONS
3.1 Survey details
This paper focuses on the review of 45 NI 34-101 technical reports of gold deposits which
were obtained from System for Electronic Document Analysis and Retrieval (SEDAR) website plus one major gold mining company’s publicly available general classification guidelines.
The purpose of the survey was to investigate the current different techniques used to classify
gold deposits into measured, indicated and inferred categories. After evaluating the different
reports, it was realized that the mineral industry players use different procedures to categorize
blocks into different classes, confirming irregular methodology used in the classification of
mineral resources. A summary table was created to include the following information as formally described in the various NI 43-101 reports: publication date, company name, location,
continent, state or province, commodity type, operation type, deposit type, interpolation
method, classification criteria, classification technique, specific gravity method, domaining
and boundary applications.
The public disclosure reports were grouped with respect to companies considered as major,
mid-tier and junior gold mining operators. The investigations covered twenty (20) junior,
twenty (20) mid-tier, and five (5) major gold mining companies in North America, South
America, Asia and Africa. Also, the general classification guidelines of a major gold mining company unlisted on SEDAR was reviewed. Tables 1–4 show details of the number of
NI 43-101 reports that were surveyed in terms of continent, country, company category,
company name, deposit type and operation type. The overall investigations were based on;
classification reports for same deposit type for same company, classification reports for same
deposit type for different companies, classification reports from different consultants on
same deposit type for different companies, and different classification estimation passes for
same mining district and mineralized belt.
After critical look at the classification guidelines used by the various companies from
SEDAR, forty-three (43) of the reports applied geometric methods and two (2) applied combined geostatistical and geometric methods. The major gold mining company unlisted on
SEDAR also applied both geostatistical and geometric techniques. The abbreviations used in
estimation is the kriging variance (KV). The KV takes into account geostatistical parameters which combine both geometric and geological inputs in characterizing the uncertainty associated with the estimated parameter. The consideration of the spatial structure
of the estimated variable and the redundancy between samples are the purposes of using
kriging variance as the criterion for classification. In the process of assigning confidence
levels, estimated blocks with high kriging variance have lower confidence than those with
lower kriging variance. Emery et al (2006) suggest the establishment of KV thresholds for
measured, indicated and inferred mineral resources classification categories based on a
given desired number of drill holes, spacing and the variogram model in a given domain.
Classifying each block into a specific measured, indicated and inferred mineral resource
category is done on the basis of KVs associated with each block and these previously
determined thresholds.
2.2.2 Classification based on 90% confidence interval
This technique involves calculating the 90% confidence intervals (CI) on the tonnage, grade
and metal content within quarterly and annual production blocks either through KV or
geostatistical conditional simulations. A recognized mining industry practice in the application of this technique is the ability to determine the drill hole spacing to be enough in
the prediction of tonnage, grade and metal content within ±15% relative precision at 90%
confidence interval within a quarterly or annual production rates. Measured resource considers quarterly production period and indicated resource corresponds to annual production
period. For the detail description of mineral resource classification based on 90% confidence
intervals, please refer to Verly et al (2014).
3 SURVEY RESULTS AND DISCUSSIONS
3.1 Survey details
This paper focuses on the review of 45 NI 34-101 technical reports of gold deposits which
were obtained from System for Electronic Document Analysis and Retrieval (SEDAR) website plus one major gold mining company’s publicly available general classification guidelines.
The purpose of the survey was to investigate the current different techniques used to classify
gold deposits into measured, indicated and inferred categories. After evaluating the different
reports, it was realized that the mineral industry players use different procedures to categorize
blocks into different classes, confirming irregular methodology used in the classification of
mineral resources. A summary table was created to include the following information as formally described in the various NI 43-101 reports: publication date, company name, location,
continent, state or province, commodity type, operation type, deposit type, interpolation
method, classification criteria, classification technique, specific gravity method, domaining
and boundary applications.
The public disclosure reports were grouped with respect to companies considered as major,
mid-tier and junior gold mining operators. The investigations covered twenty (20) junior,
twenty (20) mid-tier, and five (5) major gold mining companies in North America, South
America, Asia and Africa. Also, the general classification guidelines of a major gold mining company unlisted on SEDAR was reviewed. Tables 1–4 show details of the number of
NI 43-101 reports that were surveyed in terms of continent, country, company category,
company name, deposit type and operation type. The overall investigations were based on;
classification reports for same deposit type for same company, classification reports for same
deposit type for different companies, classification reports from different consultants on
same deposit type for different companies, and different classification estimation passes for
same mining district and mineralized belt.
After critical look at the classification guidelines used by the various companies from
SEDAR, forty-three (43) of the reports applied geometric methods and two (2) applied combined geostatistical and geometric methods. The major gold mining company unlisted on
SEDAR also applied both geostatistical and geometric techniques. The abbreviations used in