2. Harry Parker, Past Chairman, CRIRSCO
The Reliability of Resource and Reserve Classification for Public Reporting
Harry Parker, Past Chairman, CRIRSCO
Ankara, 27 January 2016

3. Introduction
The CRIRSCO codes provide some guidance, but it is left to the Competent Person to establish criteria specific to each mineral deposit being reported
With the advent of geostatistics, quantitative tools are available to forecast reliability and to establish drill spacings
This started with establishing estimation variances at the block level, and then monthly, quarterly and annual production increments
Conditional simulation is now being used to evaluate reliability on a local level and to provide assurance on individual quarterly and annual production increments and to thus check the reliability of a production schedule
In operating mines reconciliation between resource models used to declare reserves and grade control models can be used to establish reliability 3

4. What does the CRIRSCO Template Say? Inferred Mineral Resources
Quantity and grade or quality are estimated on the basis of limited geological evidence and sampling.
Geological evidence is sufficient to imply but not verify geological and grade or quality continuity.
An Inferred Resource has a lower level of confidence than that applying to an Indicated Mineral Resource and must not be converted to a Mineral Reserve. It is reasonably expected that the majority of Inferred Mineral Resources could be upgraded to Indicated Mineral Resources with continued exploration
Confidence in the estimate is usually not sufficient to allow the results of the application of technical and economic parameters to be used for detailed planning. For this reason, there is no direct link from an Inferred Resource to any category of Mineral Reserves
Caution should be exercised if this category is considered in technical and economic studies. 4

5. Inferred Mineral Resources – Comments (1 of 2)
Global outline of mineralization is possible; serves as guide to design drilling/sampling programs to convert to Indicated Mineral Resources
“Imply continuity” means mineralization controls are recognized in terms of a favorable lithology, structure, contacts etc.
CP should have knowledge of analogue deposits that permits implication of continuity
Typically resource models show neighborhood effects, weak correlation of grade between holes
Often drilling inadequate to determine width of mineralized zones
Alternative geological interpretations and ore controls are possible
May result because impossible to verify historic data locations, QAQC issues
Preliminary economic assessments and scoping studies may be inaccurate at the annual production increment scale. Thus cash flows, NPVs etc. may be unreliable. Estimates of dilution and ore loss may be unreliable; the mining method and access may change with acquisition of more information 5

6. Inferred Mineral Resources – Comments (2 of 2)
The trend over the last 10 years has been to tighten confidence required for inferred. This has been led by SME, as for example:
- An Inferred Mineral Resource can be based on interpolation between widely spaced data where there is reason to expect geological continuity of mineralization, but not excessively extrapolated from the data. The proportion of extrapolated Mineral Resource outside the nominal drill grid spacing must be limited and disclosed
- Inferred Mineral Resources should exclude material for which there are insufficient data to allow the inference of geological and grade continuity. Inferred Mineral Resources are intended to be sufficiently defined that their overall tonnages, grades and mineral contents can be estimated with a reasonable level of confidence 6

7. Example Inferred Mineral Resources from Antamina, Peru - Zinc Grades (Plan View)
Lack of drilling in favorable high-grade zinc structure; Importance of NW structural orientation missed  underestimation of tonnage and grade of copper-zinc ore on east side of lake
> 2.5% Zn, localized in structure zones
Talus prevents drilling
> 2.5% Zn
2003 Long-range Zn Model (20 x 20 m blocks)
Short-range Model 5 x 5 m blocks 7

8. Example Inferred Mineral Resources from Platreef RSA
Holes are 400 m apart
There is continuity by stratigraphic position
Limited grade continuity between holes except in 1 g/t shell based on
5 m composites 8
Source 2013 Ivanhoe Mines PEA

9. What does the CRIRSCO Template Say? Indicated Mineral Resources
An Indicated Mineral Resource is that part of a Mineral Resource for which quantity, grade or quality, densities, shape and physical characteristics are estimated with sufficient confidence to allow the application of Modifying Factors in sufficient detail to support mine planning and evaluation of the economic viability of the deposit
Geological evidence is derived from adequately detailed and reliable exploration, sampling and testing and is sufficient to assume geological and grade or quality continuity between points of observation
Mineralisation may be classified as an Indicated Mineral Resource when the nature, quality, amount and distribution of data are such as to allow confident interpretation of the geological framework and to assume continuity of mineralisation
In assessing continuity between points of observation, the Competent Person must consider the likely cut-off grade and geometric limits that would be used to prepare incremental (e.g. annual or phases) mine plans 9

10. Indicated Mineral Resources - Comments
Strong statement that Indicated Mineral Resources must support mine planning, i.e. conversion to Probable Reserves when a Pre-feasibility study is performed. This came from Canada and resulted in an upgrade of the confidence level expected for an Indicated Mineral Resource
Data quality is good; historic data verified – sometimes have to drill twin holes
Geological framework is settled
Continuity of grade, orebody boundaries must be established using cut-off grade and thickness criteria to be used for mine planning.
Typically there should be a minimum of two or three holes to define an ore pod.
Final ore/waste selection usually done by separate sampling (grade control)

11. Example Indicated and Inferred Mineral Resources from Platreef RSA
Drill spacing 100 m
High-grades follow pegmatoidal pyroxenite and harzburgite
Merensky reef analogue established 11
Source: 2013 Ivanhoe Mines

12. What Does the CRIRSCO Template Say? Measured Mineral Resources
A Measured Mineral Resource is that part of a Mineral Resource for which quantity, grade or quality, densities, shape, and physical characteristics are estimated with confidence sufficient to allow the application of Modifying Factors to support detailed mine planning and final evaluation of the economic viability of the deposit
Geological evidence is derived from detailed and reliable exploration, sampling and testing and is sufficient to confirm geological and grade or quality continuity between points of observation
 Mineralisation may be classified as a Measured Mineral Resource when the nature, quality, amount and distribution of data are such as to leave no reasonable doubt, in the opinion of the Competent Person determining the Mineral Resource, that the tonnage and grade of the mineralisation can be estimated to within close limits, and that any variation from the estimate would be unlikely to significantly affect potential economic viability
This category requires a high level of confidence in, and understanding of, the geology and the controls of the mineral deposit
Confidence in the estimate is sufficient to allow the application of technical and economic parameters and to enable an evaluation of economic viability with a high level of confidence 12

13. Measured Mineral Resource - Comments
Must support detailed mine planning
Must be able to confirm continuity between points of observation. This could mean on orthogonal sections and plans.
Must have high confidence as would be converted to a Proven Mineral Reserve in Pre-Feasibility studies, assuming high confidence also exists in the modifying factors 13

14. Example Measured Mineral Resources for Copperwood Project, Michigan USA
Source: 2010 Orvana Technical Report; grid lines spaced at 1000 ft = 300 m

15. Example Measured Mineral Resources for Copperwood Project, Michigan USA
Example cross section
Sediment hosted copper mineralization is located at the bottom of the Domino unit of the Nonesuch Formation

16. Example Measured Mineral Resources for Copperwood Project, Michigan USA
Drill Intercepts in Domino Unit and Kriged Copper Grades

17. Example Measured Mineral Resources for Copperwood Project, Michigan USA
Drill intercepts and kriged thicknesses for Domino Unit

18. What does the Template Say
What does the Template Say? Conversion of Mineral Resources to Mineral Reserves Confidence in the Modifying Factors
Modifying Factors are considerations used to convert Mineral Resources to Mineral Reserves. These include, but are not restricted to, mining, processing, metallurgical, infrastructure, economic, marketing, legal, environmental, social and governmental factors
 Measured Mineral Resources may convert to either Proved Mineral Reserves or Probable Mineral Reserves. The Competent Person may convert Measured Mineral Resources to Probable Mineral Reserves because of uncertainties associated with some or all of the Modifying Factors which are taken into account in the conversion from Mineral Resources to Mineral Reserves

19. Confidence in the Modifying Factors – The Template Provides a Checklist the CP must Consider (1 of 3)

20. Confidence in the Modifying Factors – The Template Provides a Checklist the CP must Consider (2 of 3)

21. Confidence in the Modifying Factors – The Template Provides a Checklist the CP must Consider (3 of 3)

22. Confidence in the Modifying Factors – Comments (1 of 2)
The checklist does not provide much guidance as to confidence requirements. This is left to the CP
The CP is afforded the right to convert Measured Resources to Probable reserves if there are uncertainties in the Modifying Factors
Note: In North America Proven is used instead of Proved

23. Confidence in the Modifying Factors – Comments (2 of 2)
Estimation of Mineral Reserves goes beyond the traditional brief of the mining engineer:
- The effect, if any, of natural risk, infrastructure, environmental, legal, marketing, social or governmental factors on the likely viability of a project and/or on the estimation and classification of the Mineral Reserves
- The status of titles and approvals critical to the viability of the project, such as mining leases, discharge permits, government and statutory approvals
- Environmental descriptions of anticipated liabilities. Location plans of mineral rights and titles
He must rely on colleagues to assess confidence and answer the question: is there a reasonable expectation the project will proceed?
On large projects a dozen CPs may take responsibility

24. The Template Will Likely Be Updated to Include Guidance Such as that in the 2014 SME Guide (1 of 2)

25. The Template Will Likely Be Updated to Include Guidance Such as that in the 2014 SME Guide (2 of 2)

26. What Does the Template Say? Probable Mineral Reserves
A Probable Mineral Reserve is the economically mineable part of an Indicated, and in some circumstances, a Measured Mineral Resource
The confidence in the Modifying Factors applying to a Probable Mineral Reserve is lower than that applying to a Proved Mineral Reserve
A Probable Mineral Reserve has a lower level of confidence than a Proved Mineral Reserve but is of sufficient quality to serve as the basis for a decision on the development of the deposit
A Probable Mineral Reserve derived from a Measured Mineral Resource may be converted to a Proved Mineral Reserve if the uncertainties in the Modifying Factors are removed. No amount of confidence in the Modifying Factors for conversion of a Mineral Resource to a Mineral Reserve can override the upper level of confidence that exists in the Mineral Resource. Under no circumstances can an Indicated Mineral Resource be converted directly to a Proved Mineral Reserve

27. Probable Mineral Reserves - Comment
The Template just says Probable Mineral Reserves may have lower confidence in the modifying factors compared to Proven Mineral Reserves
Some deposits will only have Probable Mineral Reserves
Most companies convert Indicated Mineral Resources to Probable Mineral Reserves without much thought as to uncertainty except for stope design, geotechnics, hydrogeology. Securities regulators are focusing increase scrutiny on infrastructure for projects in remote locations

28. What does the Template Say? Proven Mineral Reserves
A Proven Mineral Reserve is the economically mineable part of a Measured Mineral Resource. A Proved Mineral Reserve implies a high degree of confidence in the Modifying Factors
A Proven Mineral Reserve represents the highest confidence category of reserve estimate
The style of mineralisation or other factors could mean that Proven Mineral Reserves are not achievable in some deposits. Competent Persons should be aware of the consequences of declaring material of the highest confidence category before satisfying themselves that all of the relevant resource parameters and Modifying Factors have been established at a similarly high level of confidence

29. Proven Mineral Reserves - Comments
Most companies like to have the first five (sometimes four) years of Proven Mineral at the Proven level for new projects
Others require Proven Mineral Reserves to cover the payback period
Some companies do not declare Proven Mineral Reserves until a project has been successfully operated
Banks are generally happy with a mixture of Probable and Proven reserves that will cover 1.5 to 2 X the life of the loan. Repayment terms may depend on the proportion of Proven Mineral Reserves
Once a project has started there is often a reluctance to perform infill drilling to convert Probable to Proven Mineral Reserves. Reconciliation problems result and remaining reserves may be factored 29

30. Review of Best Practices in Resource Classification
From “Evolution of Mineral Resource Classification from to 2014 and Current Best Practice” by Harry Parker and Christina Dohm, Finex 2014 Julius Wernher Lecture 30

31. Continuity Very important consideration
Consider grade continuity (use variogram distance to 50% of sill for Measured, 75% of the sill for indicated, 100 to 200% of the final range for Inferred) – We consider this crude; better to inspect the cross sections and plans
In some deposits thickness is more uncertain than grade, and should govern the classification (example nickel laterite/saprolite deposits)

32. Figures from Stephenson et al. (2006)
Best Practice
Look through sections and plans
Down-grade classification where locally uncertain
Faults (Inferred corridors)
Dykes (barren or deleterious alteration e.g. talc in iron deposits)
Avoid spotted dog 32
Figures from Stephenson et al. (2006)

33. Make a Simple Conceptual Economic Analysis
Usually assume same for all years

34. Sensitivity of Economic Parameters to Changes in Grade
Small error can have serious implications on cash flows and NPV if error is a bias that persists over life-of-mine
From Parker, 2014

35. Target Relative Accuracy
Relative Accuracy = (Predicted – Actual)
Actual
Necessary Components:
- Time period or production increment
- Confidence Interval: probability within range
- Magnitude: ± %

36. Typically We Use a 90% Confidence Interval; Target Relative Accuracy ≤ 15%
Assuming a normal distribution of errors:
- Relative Accuracy is SErr/Mean
- Our target Relative Accuracy is SErr/Mean < 15%;
- SErr is standard deviation of estimation error

37. Sensitivity of Confidence Limits
If the time period is larger, SErr will be smaller because the variance of large production increments is less than the variance of small production increments. A broader drill spacing can be used for large increments to obtain the same level of accuracy
If the confidence interval is increased, say from 90 to 95%, the factor increases to The confidence limits broaden:
If the Percent Relative Accuracy broadens from say ±15 to 20%, the confidence interval broadens.

38. Guidance
Inferred: Insufficient geological information to establish confidence levels.
Indicated: ± 15% accuracy with 90% confidence over annual production increment. Actual will be within 85 and 115% of the estimate 90% of the time. Annual production increments are typically used for Pre-feasibility and Feasibility cash flows. Can stand one in year in 20 as being below 85% of the estimate – normal business risk. If actual is less than 85%, very often the mine will run a loss
Measured: ± 15% accuracy with 90% confidence over a quarterly or monthly production increment. Actual will be within 85 and 115% of the estimate 90% of the time. Quarterly or monthly production increments are typically used for Operating Budget cash flows. If error is less than 15% can usually rework the mine plan and prevent a loss
See Verly et al. (2014) for historical precedents and further explanation

39. Calculation of Simplified Confidence Limits on Production Increments
Calculate estimation variance for large blocks, which when combined can be considered to be spatially independent – May represent monthly or quarterly increments
Standard error is square root of estimation (kriging) variance
Standard error a for a period = SErr period = SErr block/√N
N is the number of large blocks in a period
90 % Confidence limits = SErr period /Mean
Try various drill spacings. Find drill spacing that yields target 90% confidence limits
Calculate estimation variance for resource model blocks using the drill spacing. Apply to block model. Helps where have irregularly spaced drilling

40. Provided courtesy of OT LLC; designs have been superseded
Oyu Tolgoi Case (2010) – Cu Grade Shells Plus Year 1 and Year 5 Production Caves
Year 5 production volume
Year 1 production volume
High -Grade Cu shell
0.6% Cu shell
Provided courtesy of OT LLC; designs have been superseded

41. Drill Spacing (actual)
Year 5
Year 1
Provided courtesy of OT LLC; designs have been superceded

42. Provided courtesy of OT LLC; designs have been superseded
Layout Using 75 x 125 m Drill Spacing; This Was Used for Preliminary Economic Assessments
Year 1 Cave
Year 5 Cave
Plan
Section
Provided courtesy of OT LLC; designs have been superseded

43. Layout Using 70 x 70 m Drill Spacing
Year 1 Cave
Year 5 Cave
Plan
Section
Provided courtesy of OT LLC; designs have been superseded

44. Layout Using 35 x 35 m Drill Spacing
Year 1 Cave
Year 5 Cave
Plan
Section
Provided courtesy of OT LLC; designs have been superseded

45. Layout Using 17.5 x 17.5 m Drill Spacing
Year 1 Cave
Year 5 Cave
Plan
Section
Provided courtesy of OT LLC; designs have been superseded

46. Confidence Limits Versus Drill Spacing
Year 1 has broader confidence intervals because it is about 10% of Year 5 tonnage
Confidence criteria on an annual basis. Measured criteria adjusted from quarterly to annual production increment

47. Current and Planned 2010 Drilling by OT LLC – 50 X 75 m Spacing
Current drilling
Designed drilling
Year 5 N
Year 5 S
Year 6 N
Year 7 N
Year 4
Year 3
Year 2
Year 1
Year 7 S
Year 6 S
100 m 47
Provided courtesy of OT LLC; designs have been superseded

48. Confidence Limits for Current Drilling and Current + Planned (2010) Drilling
First and second years are not at full production
Confidence criteria on an annual basis. Measured criteria adjusted from quarterly to annual production increment

49. Thank You 49