Consultores de Recursos Minerales S.A.

Introduction

For the past 20 years, CRM (Consultores de Recursos Minerales SA) has provided mineral resource estimation services to our clients located throughout the Americas from our headquarters in Chile. The following paragraphs briefly describe our approach to resource estimation and past experience. Additional details and references are available upon request.

Resource Estimation

Resource estimates and classification studies have been performed for numerous clients. During these studies specialized/customized statistical studies are performed, based on the observed geological controls, to determine the most appropriate estimation approach. Local grade trends and other non-stationarities are studied to ensure that the estimation is consistent with geology and the spatial distribution of grades. One example of this is our custom developed method for identifying and capping outliers based on the local behavior of grades. Following estimation, detailed visual and quantitative validations are performed to ensure that the model honors the data and the expected block grade distribution.

Resource Audits

Resource audits have been provided for numerous clients. Unlike typical audits, we do not simply focus on procedures. Where modeling or estimation issues are identified, re-estimation is performed to quantitatively define the potential impact of the issue on the resource statement.

Clients include Pan American Silver, Antofagasta Minerals, Codelco, Minera Escondida (BHP), Antamina, Teck, Glencore, Newmont, Minera Zafranal, and Resolution Copper (RTZ).

Courses

Short courses are prepared to meet the client’s requirements using previously prepared courses covering the following topics. For each topic a manual discussing basic theory and applications is available.

Basic statistics and definition of estimation domains
Spatial correlation and variability modeling
Kriging
Conditional Simulation
Resource Classification

Courses have recently been provided for Compania Minera Antamina, Antofagasta Minerals (Minera Centinela), and MMG (Las Bambas).

Special Studies

Examples of detailed studies performed to understand and optimize mining operations are:

Local variability and uncertainty models
Local anisotropy Kriging for ore control
Directional search to minimize the impact of grade trends
Indicator Kriging with conditional grade distributions
Conditional simulation of estimation unit and grade
Outlier management based on local risk
Quantitative resource classification based on uncertainty in metal above cutoff
Use of machine learning tools to improve grade control geological models
Definition of optimal ore control polygons for a polymetallic deposit with 8 ore types
Full bivariate estimation (improved co-Kriging) based on observed bivariate distribution
Dilution and blast heave modeling

Personnel and Corporate Experience

Jeff Sullivan PhD, PE, RG, FAusIMM – Over 30 years experience in resource estimation and geostatistical modeling. Doctorate in geostatistics (Stanford) where he developed the Probability Kriging estimation method. Has directed work at CRM for the last 20 years. Previously worked in project development at BHP Minerals.

Silvia Satchwell, FAusIMM - Over 30 years experience in resource estimation and geostatistical modeling. Responsible for geological modeling and reviews at CRM (20 years). Previous experience at BHP Minerals, Manville, Southern Peru Copper, and Roan Consolidated Mines.

past Projects

Evaluation of the Value of Additional Drilling for the MM Project, Chile

The MM orebody contains high grade copper, silver and arsenic breccias surrounded by copper bearing stockworks. A resource model for the entire deposit was created. In the portion of the deposit that would be produced first, additional drilling is proposed. To determine the value of this drilling in terms of reducing project risk, multiple conditional simulation of the deposit before and after the proposed drilling were prepared. Applying the mining plan to the simulations, the uncertainty in the annual tonnage of copper produced and therefore the uncertainty in the annual cash flow were developed. Given the uncertainty in the cash flows, the uncertainty in the project NPV could be computed both before and after the additional drilling. Given these distributions the risk premium (difference in discount rate) associated with developing the project without additional drilling was determined and converted to a dollar value to define the value of the drilling.

Evaluation of the Uncertainty in Annual Production, Esmeralda Zone, El Teniente

To define the potential annual fluctuations in head grade from the El Teniente block caving operation, multiple simulations of the lithology and grade were prepared. The production sequence was then superimposed on the simulations. Each simulation was extracted independently and the uncertainty in grade was computed.

Uncertainty in the Tonnage of Exotic Mineralization, Extencion Norte de Mina Sur

High grade exotic mineralization is localized by paleochannels. A conditional simulation was used to define the uncertainty in the locations of the surfaces and the location of the lateral contacts.

Accounting for Peak Concentrations in an Oxide Deposit, Damiana Deposit, El Salvador

Copper grades in the exotic Damiana deposit peak strongly along a dipping surface associated with the presence of pods of green and black copper minerals. Preservation of the vertical trends associated with this peak behavior was critical to defining the resource. Use of a standard estimator with any vertical search would cause underestimation of grades at the peak and overestimation of grades on the “shoulders”. The best results (superior to universal Kriging) were found using a two step Kriging method. First the deposit was unfolded so that the peak grades were found at a single level. Next grades were estimated using ordinary Kriging with a flat horizontal search. In the second Kriging step, a 3D search was used in a simple Kriging estimate. The local mean for the simple Kriging was the estimated grade from the first pass (horizontal) estimate. This approach accurately reproduced the peak grades.

CRM-sa