---

---

Exhibit 99.1

 

GOLD RESERVE INC.

TECHNICAL REPORT ON THE SIEMBRA MINERA PROJECT, BOLIVAR STATE, VENEZUELA

NI 43-101 Report

Qualified Persons:

Richard J. Lambert, P.E., P.Eng. Hugo Miranda, C.P.

José Texidor Carlsson, P.Geo. Kathleen A. Altman, Ph.D., P.E. Grant A. Malensek, P.Eng.

March 16, 2018

RPA Inc. 55 University Ave. Suite 501 I Toronto, ON, Canada M5J 2H7 I T + 1 (416) 947 0907 www.rpacan.com

---

 

Report Distribution

Name

No. of Copies

Client

RPA Filing

1 (project box)

Roscoe Postle Associates Inc.
55 University Avenue, Suite 501
Toronto, ON M5J 2H7
Canada
Tel: +1 416 947 0907
Fax: +1 416 947 0395
mining@rpacan.com

 

---

www.rpacan.com

 

 

---

 

20 ENVIRONMENTAL STUDIES, PERMITTING, AND SOCIAL OR COMMUNITY IMPACT

20-1

 

 

---

www.rpacan.com

 

 

---

 

 

---

www.	rpacan.com

 

           

 

 

---

www.	rpacan.com

 

 

---

www.rpacan.com

1 SUMMARY

EXECUTIVE SUMMARY

Roscoe Postle Associates Inc. (RPA) was retained by Gold Reserve Inc.(GRI), and its wholly owned subsidiary GR Engineering Barbados, Inc. (GRE) to prepare an independent Technical Report on the Siembra Minera Project (the Project), located in Bolivar State, Venezuela. The operating company Empresa Mixta Ecosocialista Siembra Minera, S.A. (Siembra Minera), which holds the rights to the Siembra Minera Project, is a mixed capital company with 55% being owned by a Venezuelan state entity [owned by the Bolivarian Republic of Venezuela through the Corporación Venezolana de Minería (CVM)] and 45% by GR Mining Barbados, Inc. (GRM), a wholly-owned subsidiary of GRI. GRE has been set up to perform engineering, procurement, construction, and operation of the Project.

The Project is a combination of the Brisas and Cristinas properties into a single project now called the Siembra Minera Project. The purpose of this report is to provide GRI and GRE with an initial assessment of the Siembra Minera Project including a resource estimate, conceptual mine plan, and a preliminary economic review. This Technical Report conforms to NI 43-101 Standards of Disclosure for Mineral Projects. RPA visited the Project on September 19, 2017.

The Siembra Minera Project is a gold-copper deposit located in the Kilometre 88 mining district of Bolivar State in southeast Venezuela. Local owners and illegal miners have worked the property for many years. Shallow pitting and hydraulic methods were used to mine the upper saprolite zone, and coarse gold was recovered by gravity concentration and amalgamation with mercury. Most of the large-scale exploration work at Cristinas was performed by Placer Dome Inc. (Placer), which worked on the property from 1991 to 2001. At Brisas, GRI carried out the exploration program on the concession from 1992 to 2005. The most recent Technical Report for Cristinas is dated November 7, 2007, which is based on a feasibility study and includes historic mineral reserves. The most recent Technical Report for Brisas is dated March 31, 2008, which is also based on a feasibility study and includes historic mineral reserves.

RPA has relied on data derived from work completed by previous owners on the Cristinas concessions and by GRI on the Brisas concessions. The current resources for Cristinas were estimated by RPA based on the drill hole data supplied by Corporación Venezolana de

 

---

www.rpacan.com

Guayana (CVG) to GRI in 2002. The database had 1,174 drill holes and 108 trenches which were included in the Cristinas database. Hard copies of the assay data sheets were not available, however, GEOLOG data files from Placer were provided including assay data, geological descriptions, structural data, geotechnical data, and check sample data. The current resources for Brisas were estimated by RPA based on drill hole data supplied by GRI in Geovia GEMS format which formed the basis of the last Technical Report by Pincock Allen & Holt (PAH) in 2008.

This report is considered by RPA to meet the requirements of a Preliminary Economic Assessment (PEA) as defined in Canadian NI 43-101 regulations. The mine plan and economic analysis contained in this Technical Report are based, in part, on Inferred Mineral Resources, and are preliminary in nature. Inferred Mineral Resources are considered too geologically speculative to have mining and economic considerations applied to them that would enable them to be categorized as Mineral Reserves. There is no certainty that economic forecasts on which this PEA is based will be realized.

CONCLUSIONS

RPA offers the following conclusions by area.

GEOLOGY AND MINERAL RESOURCES

• A number of exploration programs completed by Placer and GRI were successful in locating and defining the extents of the various mineralized zones on each of their respective property holdings. The recently established Siembra Minera Economic Zone has unified the land tenure.
• The geology of the deposit is well understood in general. RPA is of the opinion that the distribution of high grade areas in the Main Zone should be studied in more detail.
  In the southern two-thirds of the Cristinas concessions and the entirety of the Brisas concessions, the mineralization occurs in a large tabular body, which strikes approximately north-south and dips moderately to the west. In the northern third of the Cristinas concessions, the mineralization can occur as pipe-shaped forms, and as thinner tabular forms with sub-vertical dips and strikes to the southeast.
• The large tabular, strataform mineralized zone (referred to herein as the Main Zone) forms most of the Mineral Resource. The Main Zone has a minimum thickness of 10 m at the south end and reaches a maximum thickness of 350 m. The average thickness is approximately 200 m. While the southern limits of the Main Zone have been outlined by the existing drilling pattern with a reasonable degree of confidence, the down-dip limits have not been defined by drilling. The northern limits of the Main Zone are also reasonably well defined by the existing drilling pattern.

 

---

www.rpacan.com

• The drill hole information collected by Placer and GRI was merged into one master database that was then used to prepare the Mineral Resource estimate. Additional drill hole information collected by Crystallex International Corporation (Crystallex) on the Cristinas concessions could not be used to prepare the current estimate of the Mineral Resources, as the detailed information required was not available. The drill hole data from Placer contained drilling information and analytical results up to 1997 while the drill hole data from GRI included information up to 2006.
• In RPA’s opinion, the drill hole data is adequate for use in the preparation of Mineral Resource estimates.
• The outline of the gold mineralization was created by drawing wireframes using approximately a 0.20 g/t Au cut-off grade and the copper mineralization was outlined using broad wireframes based on approximately a 0.04% Cu cut-off grade. A total of 24 wireframes were constructed to represent the gold mineralization zones and six wireframes to represent the copper mineralization zones. RPA also prepared wireframe surfaces to represent the three main weathering profiles for the mineralized zones: oxide saprolite, sulphide saprolite, and hard rock.
• RPA applied variable capping values for gold and copper grades for each of the mineralized wireframe domains. The capped assay values were composited into three metre lengths. The composites were then used to estimate the gold and copper grades into a grade-block model that used block sizes of 10 m by 10 m by 6 m. Gold and copper grades were estimated into blocks using inverse distance squared and dynamic anisotropy with the Surpac v.6.8 software package. The estimated gold and copper grades were used to calculate Net Smelter Return (NSR) values for each mineralized block.
• Mineral Resources were prepared using an NSR cut-off value of US$7.20/t for the oxide saprolite and US$5.00/t for the sulphide saprolite and fresh rock. An open pit shell was created using the Whittle software package to constrain reporting of the Mineral Resources.
• The Mineral Resource estimate conforms to Canadian Institute of Mining, Metallurgy and Petroleum (CIM) Definition Standards for Mineral Resources and Mineral Reserves dated May 10, 2014 (CIM, 2014).
• The Mineral Resources are estimated at 10 million tonnes at an average grade of 1.02 g/t Au and 0.18% Cu containing 318,000 ounces of gold and 17,000 tonnes of copper in the Measured category, 1.17 billion tonnes at an average grade of 0.70 g/t Au and 0.10% Cu containing 26.5 million ounces of gold and 1.2 million tonnes of copper in the Indicated category. Mineral Resources in the Inferred category are estimated at 1.30 billion tonnes at an average grade of 0.61 g/t Au and 0.08% Cu containing 25.4 million ounces of gold and 1.0 million tonnes of copper.

MINING

• Mine production is scheduled to be carried out at a maximum mining rate ranging from 330 ktpd to 380 ktpd of total material.
• Stripping ratios are expected to average 1.16 over the Life of Mine (LoM) plan

 

---

www.rpacan.com

• A separate equipment fleet of smaller excavators and articulated dump trucks is included in the mining capital for saprolite mining in the first 10 years. Typically, undisturbed saprolite material can be difficult to mine as the moisture creates operation problems. As the Project area has essentially been disturbed, RPA has assumed most saprolite is handled by the larger equipment fleet. The larger mine fleet is more productive and prior experience at Cristinas shows that rigid frame trucks can operate in the saprolite.
• Stockpiles are required for blending the process feed to achieve sufficient copper grades in flotation to produce a copper concentrate above 20%. Stockpiles fluctuate year to year, but achieve maximum capacity of just over 70 million tonnes.

MINERAL PROCESSING

• Both Brisas and Cristinas were developed to the feasibility-level stage and beyond in 2006 to 2007 so the quantity of information available is greater than would typically be available at the PEA stage of a project.
• The material to be mined from Siembra Minera is demonstrated to be amenable to both cyanide leaching and to sulphide flotation. For materials that contain lower concentrations of copper, cyanide leaching is more cost effective and for material that contains higher concentrations of copper, sulphide flotation is more cost effective.
• The prior metallurgical test work met industry standards at the time the studies were completed, however, technology has progressed in the subsequent ten plus years and industry standards have evolved. Current standards include testing of a large number of variability samples and development of geometallurgical models, as opposed to testing composite samples to represent “average” material to be processed, which was the emphasis for the Brisas test program.

ENVIRONMENT

• GRE is in the process of preparing environmental reports and programs to meet municipal, provincial, and national regulatory requirements, as well as generally accepted international standards.
• Two separate but parallel Environmental and Social Impact Assessments (ESIA) are being prepared for the Project, one that meets Venezuelan regulatory requirements and one that meets international standards and guidelines.
• A conceptual plan for small-scale mining management is in place. The conceptual plan includes relocation of the artisanal miners away from the active, large scale mining operations and establishment of an oxide saprolite processing and stockpile area with concrete tailings ponds that collect and transport tailings from the artisanal mining operations to the Project tailings management facility (TMF).

RECOMMENDATIONS

Given the positive economic results presented in this report, RPA recommends that the Project be advanced to the next stage of engineering study and permitting.

 

---

www.rpacan.com

RPA offers the following recommendations.

GEOLOGY AND MINERAL RESOURCES

	Acquire new topographic data.
	Drill approximately 150 to 200 drill holes totalling approximately 75 km to 100 km. This drilling would have a number of objectives including:
	o	Conversion of Inferred Mineral Resources to Indicated with priority set on Inferred Mineral Resources situated in the 5 and 10-year pit shells.
	o	Drilling to determine the extent of mineralization at depth in the Main Zone as this will determine the limits of the largest possible pit and help with the location of features such as dumps and roads.
	o	Better definition of the copper mineralization in the Main Zone footwall.
	o	Improving preliminary artisanal mining sterilization assumptions.
	o	Condemnation drilling of proposed waste rock storage sites.
	o	Closer spaced drilling in the El Potaso area between Brisas and Cristinas concession areas.
	o	Drilling on the northwest extensions of the mineralization in the Morrocoy and Cordova areas.
	o	Drilling on the Cristinas Main Zone for density measurements.
	Improve understanding of the geological and structural controls on the shapes and local trends of high grade lenses in the Main Zone. Northwest striking cross-faults need to be identified and modelled and structural sub-domains built to improve future variography studies and dynamic anisotropy trend surfaces. This will improve the local accuracy of future gold and copper grade models.
	Carry out additional 3D mineralization trend analysis studies, domain modelling, and variography work for the gold and copper mineralization. This will also assist in evaluating if additional 5-spot drill holes are needed to support the Indicated classification in some areas with more complex geology.
	Depending on the outcome of new variography work, build gold and copper models
	using	ordinary kriging.
	Develop a new lithology model once new drill holes have been drilled so that an improved material densities model can be created.
	Build a structural model.
	For the proposed drilling, implement field and coarse duplicate sampling programs at Siembra Minera at a rate of approximately 1 in 50.
	Acquire three or four matrix matched certified reference materials that approximate the cut-off grade, average grade, and high grades and insert them in all future drill programs at the Project at a rate of approximately 1 in 25.
	Implement external laboratory check assays at a rate of approximately 1 in 20.

 

---

www.rpacan.com

MINING

• RPA is of the opinion that one of the most important factors influencing mining will be the amount of water entering the pit. RPA recommends contracting a groundwater hydrologist to evaluate the combined Project based on past work.
• A LoM schedule should be generated for the mining and processing of the Siembra Minera mineralized material. This study should include optimization and blending of the materials to achieve a sufficiently high copper grade to produce a copper concentrate grade above 20%.
• A trade-off study should be completed for the backfilling of the open pit with waste rock and/or neutralized tailings.
• A geotechnical investigation program should be carried out to confirm the subsurface conditions under the proposed new open pit, waste dump locations, and stability analysis undertaken to verify design recommendations.

MINERAL PROCESSING

	Every effort should be made to acquire access to the detailed metallurgical and plant
	data	for Cristinas. In the absence of that data, detailed metallurgical sampling and
	testing are required to provide the information required to design the oxide leaching
	plant.
	Additional test work should be conducted for the flotation plant using variability samples
	taken	from throughout the deposits with particular emphasis on Cristinas where limited
	variability testing was done using the flotation flowsheet. Currently, industry standard emphasizes the use of variability samples as opposed to the composite samples that
	were	predominantly used in previous flotation testing.
	RPA is of the opinion that there is considerable potential for optimization of the flowsheet of the Siembra Minera Project. Examples include:
	o	Increased efficiency if larger equipment sizes are utilized in the design. Due to cost savings and enhanced performance, the sizes for grinding mills and flotation cells have increased substantially. As examples, semi-autogenous grinding (SAG) mills that are now available are as large as 12.2 m diameter by 8.8 m long as opposed to the 11.6 m by 6.7 m that are in the current design and flotation cells now have capacities of 600 m3 instead of the 160 m3 that are in the current design. The larger pieces of equipment result in a reduced footprint and fewer pieces of equipment and, therefore, lower installed costs.
	o	The use of an adsorption desorption recovery (ADR) that is designed for the combined Project will probably result in less cost than merely doubling the size of the current design. In addition to this, consolidating the ADR from the oxide leach plant into a plant that can later be expanded to process the doré from the flotation plant has the potential to not only cut costs but also reduce security concerns and efforts.
	RPA is of the opinion that the current conceptual design for the oxide leach plant does not include the best options for Siembra Minera. Areas that require detailed evaluations include:

 

---

www.rpacan.com

	o	Use of carbon-in-leach (CIL) instead of carbon-in-pulp (CIP) particularly since the plant designs for both Cristinas and Brisas were changed to CIL from CIP during previous studies.
	o	Investigate elimination of the copper circuits. Data from the Cristinas feasibility
		study	shows that copper is only soluble in the sulphide saprolite and that it is
		not soluble in material that has lower copper concentrations. Therefore, the copper circuit should not be needed as the sulphide saprolite that contains higher concentrations of copper will be processed in the flotation plant and not in the oxide leach plant.
	o	Changes to the gravity separation circuit. The use of continuous centrifugal concentrators instead of batch units to eliminate manual labour and reduce potential for theft. Use intensive cyanide leaching to process the gravity gold concentrate instead of shaking tables. Prior studies showed that intensive cyanide leaching was preferable for treatment of the gravity concentrate for both Brisas and Cristinas.
	o	Selection of designs that are appropriate for processing clay-like saprolitic material, including:
		§	Appropriate tank sizing using slurry densities that are consistent with the material that has a low specific gravity and is viscous in nature
		§	Proper agitator selection
		§	Selection of pumps and design of piping
	Design of the TMF for the combined Project is preliminary. Further detailed geotechnical work is required to complete a design for the final tailings. Preliminary
	plans	are to use the feasibility level design from the SNC-Lavalin 2007 study as Stage
	1 of construction with the final tailings inundating the Stage 1 structure.

ENVIRONMENT

• GRI has held discussions with the small miners, indigenous groups, and local people. RPA recommends continuing discussions with these groups.
• Due to the increase in mineral resources, additional work is required for the increased waste rock dump (WRD) and TMF, and redesign/update of the acid rock drainage (ARD) mitigation measures.
• A new ESIA will be required for the combined project with an updated project plan and in conjunction with detail design and feasibility study.

COSTS AND ECONOMICS

• After the designs are complete for the Siembra Minera Project, a new capital and operating cost estimate should be completed.
• An updated copper concentrate marketing study should be completed. Recent changes in the world copper concentrate supply have reduced treatment and refining charges for copper and reduced participation charges.

PROPOSED PROGRAM AND BUDGET

RPA’s proposed program for the next stage of study is summarized in Table 1-1.

 

---

www.rpacan.com

TABLE 1-1 PROPOSED PROGRAM

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

ECONOMIC ANALYSIS

The economic analysis contained in this report is based, in part, on Inferred Mineral Resources, and is preliminary in nature. Inferred Mineral Resources are considered too geologically speculative to have mining and economic considerations applied to them that would enable them to be categorized as Mineral Reserves. There is no certainty that economic forecasts on which this PEA is based will be realized.

A Cash Flow Projection has been generated from the LoM production schedule and capital and operating cost estimates, and is summarized in Table 1-4. All currency is in US dollars (US$ or $). A summary of the key criteria is provided below.

ECONOMIC CRITERIA

PRODUCTION

• The LoM production plan assumes that leach plant detailed engineering/early earthworks will commence in Q1 of Year -2.
• The LoM production plan assumes concentrator plant detailed engineering will commence in Q1 of Year -2.
• A 2-year pre-production period for the leach plant, 2 additional years for completion of the flotation concentrator, and a 45 year overall mine life.
• The leach plant has nameplate capacity of 15,000 tpd from year 1 through year 10, which increases in year 11 to 35,000 tpd through year 45 (End of Mine, or EoM) (5.8 Mtpa to 12.25 Mtpa, respectively).
• The concentrator plant has nameplate capacity of 140,000 tpd from year 3 through year 10, which decreases in year 11 to 105,000 tpd through year 45 EoM (58 Mtpa to 36.75 Mtpa, respectively).

 

---

www.rpacan.com

• Total combined leach and concentrator production is 2.0 billion tonnes, at a grade of 0.70 g/t Au, 0.50 g/t Ag, and 0.090% Cu.
• The copper head grades in the mine plan are 302 Mt at 0.017% Cu and 1,703 Mt at 0.106% Cu for the leach and concentrator plants, respectively. However, the leach plant does not recover copper, thus the overall average copper head grade in the total mill feed is 2,005 Mt at 0.090% Cu.
• Average overall metal recovery of 84% Au, 53% Ag, and 84% Cu.
• Total recovered metal of 38.1 Moz Au, 17.1 Moz Ag, and 3.3 billion lb Cu.
• Average LoM annual recovered metal production of 847 koz Au, 380 koz Ag, and 78 million lb Cu.
• Average annual recovered metal production in Years 3 through 18 of 1,229 koz Au, 469 koz Ag, and 77 million lb Cu.
• Average annual recovered metal production in Years 19 through 45 EoM of 674 koz Au, 353 koz Ag, and 78 million lb Cu.

REVENUE

	Doré payable factors at refinery are 99.9% Au and 98% Ag.
	Copper concentrate average payable factors at smelter are 98% Au, 97% Ag, and 95.8% Cu.
	Payable metal sales for the Project total 37.6 Moz Au, 16.6 Moz Ag, and 3.2 billion lb Cu split as follows:
	o	From Doré: 14.4 Moz Au and 4.1 Moz Ag.
	o	From Concentrate: 23.2 Moz Au, 12.5 Moz Ag, and 3.2 billion lb Cu.
	Metal prices: US$1,300 per troy ounce Au; US$17 per troy ounce Ag and US$3.00 per pound Cu.
	NSR for doré includes transport and refining costs of $0.50 per ounce doré and $6 per ounce gold/$0.40 per ounce silver, respectively.
	NSR for copper concentrate includes:
	o	Cost Insurance and Freight (CIF) charge of $103 per wet tonne concentrate
		(8%	moisture content) consisting of:
		§	Road Transport (350 km one way): $11/t
		§	Port Charges (Puerto Ordaz) : $17/t
		§	Ocean Transport (Europe): $75/t.
	o	Smelter treatment charge of $95 per dry tonne concentrate.
	o	Smelter refining charges of $0.095/lb Cu, $6/oz Au, and $0.40/oz Ag.
	o	Copper price participation is not included.

 

---

www.rpacan.com

COSTS

• Pre-production period to CIP plant First Production: 24 months (January Year -2 to December Year -1).
• Pre-production period to concentrator First Production: 48 months (January Year -2 to December Year 2).
• Project development capital totals $2.57 billion, including $459 million in contingency (22% of direct and indirect capital).
• Sustaining capital of $1.42 billion.
• Average unit operating costs in $/t milled over the mine life:

 

ROYALTIES AND GOVERNMENT PAYMENTS

Royalties and other government payments total $5.6 billion, or $2.77/t milled, over the LoM as shown in Table 1-2.

TABLE 1-2 ROYALTIES AND GOVERNMENT PAYMENTS

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

The Project will pay an annual NSR royalty to Venezuela on the sale of gold, copper, and silver and any other strategic minerals of 5% for the first ten years of commercial production and 6% thereafter.

The Project is subject to an additional 3% NSR annual royalty called Special Advantages Tax which is a national social welfare fund.

 

---

www.rpacan.com

The Project is subject to a 1% gross revenue levy as part of the Science, Technology and Innovation Contributions fund (LOCTI).

Customs duties and Value Added Taxes (VAT) are assumed to be waived for the Project.

INCOME TAXES, WORKING CAPITAL, AND OTHER

Income taxes/contributions, upfront working capital, and reclamation/closure costs total $8.3 billion as shown in Table 1-3. Withholding taxes on corporate dividends and interest payments are not incorporated into the Project economic analysis.

TABLE 1-3 INCOME TAXES, WORKING CAPITAL, AND OTHER

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

Anti-drug and Sport Contributions

These profit-based taxes are assessed at 1% of current year and previous year operating income, respectively. The annual operating margin is calculated by taking annual gross revenues and deducting all operating costs and depreciation/amortization allowances.

Corporate Income Tax

The Project economic analysis incorporates a sliding scale of tax rates applicable on income based on Project phases starting in Year 1 of commercial production as follows:

• Years 1 through 5: 14%
• Years 6 through 10: 19%
• Years 11 through 15: 24%
• Years 16 through 20: 29%
• Years 21+: 34%

Year 1 is the first year of gold production, after commissioning of the 15,000 tpd oxide plant.

 

---

www.rpacan.com

Deductions from income for the purpose of estimating income subject to tax include the following items:

	Operating Expense
	Expensed operating costs are deducted 100% in year incurred.
	Stockpile adjustments
	As a result of large stockpiles of mill feed being generated during the life of the mine, the Project economic analysis includes annual adjustments to EBITDA to match mining
	costs	with recognized revenue. The net effect of these adjustments over the life of the
	mine	is zero but the adjustments increase EBITDA in years where stockpiling exceeds
	processing and inversely decrease EBITDA when processing stockpile material exceeds stockpile placement amounts.
	Depreciation/Amortization
	o	All prior expenditures before January 2018 are considered sunk with respect to this analysis.
	o	Depreciation commences once the facilities are placed into service and the mine and mill are operating.
	o	Heavy mine fleet equipment capital is depreciated using 8-year straight line (SL) method. Light vehicle capital is depreciated using 5-year SL method.
	o	All process and infrastructure capital are depreciated using the Units of Production (UoP) method.
	o	Capitalized pre-production activities such as pre-stripping and water management are amortized the UoP method.
	o	The Project economic analysis incorporates an accelerated depreciation methodology which combines the first 12 years of annual SL depreciation allowances with the standard UoP cost basis. The resulting combined UoP/SL basis is then re-calculated using the UoP method. After 12 years, the depreciation allowances come directly from each UoP or SL category.
	o	Reclamation costs are amortized during the LoM by an annual accrual of $0.035/t mined ($150 million cost divided by 4.33 billion tonnes mined). This allowance is adjusted annually by periodic reclamation capital expenditures during the LoM.
	Other Deductions
	Other	deductions from income for the purposes of estimating taxable income include
	management fees which amount to 5% of annual operating and capital costs. The annual management fees derived from operating costs are within the G&A opex category and thus expensed 100% in the year incurred while the annual fees derived
	from	capital costs are amortized using the UoP method starting in the year they are
	incurred.
	Loss Carryforwards
	Income tax losses may be carried forward indefinitely but may not be used for prior tax years.

 

---

www.rpacan.com

Upfront Working Capital

A total of $195 million has been allocated for upfront working capital in Years 1 to 4. This amount covers year over year changes in accounts receivable and payable plus consumable inventory.

Reclamation/Closure Costs

The Project economic analysis has a $150 million LoM closure cost estimate.

Salvage

No salvage value was estimated as part of the Project economic analysis.

CASH FLOW ANALYSIS

The Project as currently designed has significant variations in the mining schedule, processing methods, and head grades over its planned 45-year life. These variations are shown in Figures 1-1 and 1-2 and the resulting impact on the pre-tax free cash flow profile is shown in Figure 1-3.

FIGURE 1-1

MINE VS. MILL PRODUCTION

 

---

www.rpacan.com

FIGURE 1-2 MILL PRODUCTION PROFILE BY PLANT

FIGURE 1-3 PROJECT PRE-TAX METRICS SUMMARY

Table 1-4 shows the LoM total metrics for the Project as currently designed. Due to the length of the 45-year mine life, the full annual cash flow model is presented in Appendix 1.

 

---

www.rpacan.com

TABLE 1-4 INDICATIVE PROJECT ECONOMICS

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

 

---

www.rpacan.com

On a pre-tax basis, the undiscounted cash flow totals $28,164 million over the mine life. The pre-tax Internal Rate of Return (IRR) is 36.8%, and simple payback from start of commercial production occurs in 3.8 years. The pre-tax Net Present Values (NPV) are:

• $11,209 million at a 5% discount rate.
• $5,534 million at a 10% discount rate.

On an after-tax basis, the undiscounted cash flow totals $20,223 million over the mine life, the IRR is 31.1%, and simple payback from start of commercial production occurs in 4.1 years. The after-tax NPVs are:

• $8,101 million at a 5% discount rate.
• $3,930 million at a 10% discount rate.

The average annual gold sales during the forty-five years of operation is 836 koz per year (37.6 Moz over the LoM) at an average all in sustaining cost (AISC) of US$483 per ounce. Table 1-5 shows the AISC build up which is net of a US$262/oz copper and silver by-product credit (nbp).

TABLE 1-5 ALL-IN SUSTAINING COSTS COMPOSITION

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

 

---

 

Figure 1-4 shows the annual AISC trend during the mine operations against an overall average AISC of US$483/payable oz over the 45-year LoM at an annual production rate of 836 koz Au per year. The AISC variations are mainly driven changes in grades, mine schedule, and processing methods. The AISC metric can range from US$309/oz to US$992/oz Au in a given year (excluding final year spike in Year 45 of $1,956/oz) but can be subdivided into three distinct phases:

• Phase 1: Years 1 and 2 (CIP only) - 133 koz/yr Au at $853/oz.
• Phase 2: Years 3 through 18 (mining highest grades) - 1,191 koz/yr Au at $411/oz.
• Phase 3: Years 19 through 45 EoM (mining lower grades) - 665 koz/yr Au at $554/oz.

FIGURE 1-4

ANNUAL AISC CURVE PROFILE

 

---

www.rpacan.com

SENSITIVITY ANALYSIS

Project risks can be identified in both economic and non-economic terms. Key economic risks were examined by running cash flow sensitivities:

• Head grade
• Gold recovery
• Gold price
• Operating costs
• Capital costs
• Discount rates

Pre-tax NPV and IRR sensitivities over the base case has been calculated for -20% to +20% variations metal-related categories. For operating costs and capital costs, the sensitivities over the base case has been calculated at -15% to +35% variation. The sensitivities are shown in Table 1-6 and in Figures 1-5 and 1-6, respectively.

 

---

www.rpacan.com

TABLE 1-6 PRE-TAX SENSITIVITY ANALYSIS

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

 

---

www.rpacan.com

FIGURE 1-5 PRE-TAX NPV 10% SENSITIVITY ANALYSIS

FIGURE 1-6 PRE-TAX IRR SENSITIVITY ANALYSIS

A sensitivity analysis of discount rates is presented in Figure 1-7 and 1-8 and shows that the Project as currently designed would be NPV positive through a 20% discount rate.

 

---

www.rpacan.com

FIGURE 1-7 PRE-TAX DISCOUNT RATE SENSITIVITY ANALYSIS

FIGURE 1-8 AFTER-TAX DISCOUNT RATE SENSITIVITY ANALYSIS

 

---

www.rpacan.com

TECHNICAL SUMMARY

PROPERTY DESCRIPTION AND LOCATION

The Siembra Minera Project is located in the Kilometre 88 mining district of Bolivar State, in southeast Venezuela at Latitude 6° 11’ North and Longitude 61° 28’ West. The property is approximately 3.5 km west of Highway 10. Las Claritas is the closest town to the property.

The Project site is located in the Guyana region, which covers approximately one-third of Venezuela’s national territory. The closest nearby large city is Ciudad Guayana, with approximately 1,050,000 inhabitants (2001), situated on the Orinoco River near its confluence with the Caroní River. Ciudad Guayana consists of the old town of San Félix to the east and the new town of Puerto Ordaz to the west. Puerto Ordaz is home to most of the major industrial facilities such as aluminum smelters and port facilities. Puerto Ordaz has major port facilities accessible to ocean-going vessels from the Atlantic Ocean via the Orinoco River, a distance of approximately 200 km. There is regularly scheduled commercial airline service to Puerto Ordaz from Caracas.

Highway 10 provides paved access from Ciudad Guayana, which is 373 km northwest of the property, to within 3.5 km of the Project site. Unpaved roads provide the remaining 3.5 km of access.

The Project area encompasses approximately 18,951 ha and has been designated as an Economic Zone by the Venezuelan Government.

HISTORY

Gold in the Siembra Minera region was first discovered in 1920. Gold mining in the Project area was initiated in the 1930s and continued sporadically on a minor scale until the early 1980s when a gold rush occurred. Approximately 5,000 to 7,000 small miners worked alluvial and saprolite-hosted gold deposits using hydraulic mining techniques. The amount of gold recovered is unknown and much of the area of the concessions is now covered with tailings.

Placer conducted essentially all of the modern exploration on Cristinas during its tenure on the property from 1991 to 2001. Placer completed line cutting, mapping, rock and soil sampling, geophysics, and drilling of 1,174 drill holes for a total of 158,738 m of drilling. In 2003, Crystallex undertook drilling of 12 holes totalling 2,199 m to confirm the tenor of mineralization

 

---

www.rpacan.com

presented in the pre-existing database and also assayed check samples. Between 2003 and 2007, Crystallex released at least two feasibility studies and several resource and reserve estimates for Cristinas, all of which are historic in nature and should not be relied upon.

The Brisas concession was acquired by GRI in August 1992 with the acquisition of Compañia Aurifera Brisas del Cuyuni C.A. A large stratabound gold-copper mineralization was discovered in both alluvial and hard rock material by a drilling program in 1993. A majority of the exploration and development drilling took place in 1996 and 1997, with additional drilling completed in 1999, 2003, 2004, and 2005. As of 2005, 802 exploration holes had been drilled including 186,094 m of core drilling and 189,985 m of exploration core and auger drilling. In 2005-2006, an additional 76 holes were drilled on the Brisas concessions for geotechnical and other studies. A number of resource estimates have been completed for the Brisas deposit, all of which are superseded by the current Mineral Resource estimate in this report. A pre-feasibility study was carried out in 1998 and a feasibility study in 2005, with a feasibility update in 2008, all including historic reserve estimates.

GEOLOGY

The Siembra Minera Project is within the Guyana Shield in northern South America. The shield covers easternmost Colombia, southeastern Venezuela, Guyana, Suriname, French Guiana, and northeastern Brazil. The Venezuelan portion of the shield is subdivided into five geological provinces with different petrological, structural and metallogenic characteristics. The provinces are, from oldest to youngest, Imataca, Pastora, Cuchivero, Roraima, and Parguaza. Only the Imataca, Pastora and Roraima provinces are found in the vicinity of the Siembra Minera deposit.

The Siembra Minera deposit lies within a portion of the lower Caballape Formation volcanic and volcanic-related sedimentary rocks. The units present are (1) andesitic to rhyolitic tuffaceous volcanic beds, (2) related sedimentary beds, and (3) a tonalitic intrusive body. All rocks have been tilted and subjected to lower greenschist facies metamorphism. It is thought, based on information from nearby properties, that the Siembra Minera Project occupies one limb of a large regional fold. Limited direction-indicating structures show the strata to be top-up. In the main mineralized trend, moderate to strong foliation is oriented N10°E and dipping 30° to 55° northwest. This foliation appears to be parallel to the original bedding and tends to

 

---

www.rpacan.com

be strongest in the finer-grained rocks. A much weaker foliation orientation appears in outcrop exposures, striking north-northwest and dipping to the southwest.

There are four distinct types of gold and copper mineralization present at Brisas, defined by geometry, associated minerals, and the gold-copper ratio. These zones are the Blue Whale body, disseminated gold+pyrite (± copper), disseminated high copper, and shear-hosted gold. Only the first three types are encountered within the proposed pit geometry.

Two distinct styles of mineralization are present at Cristinas: hydrothermal breccia-hosted mineralization at Mesones-Sofia and stratiform mineralization at Conductora, Morrocoy, and Cordova. The vast majority (approximately 95%) of the gold at Cristinas is contained in the stratiform mineralized zone.

EXPLORATION STATUS

Drilling at Brisas was carried out by GRI from late 1992 to 2006 and consisted of 975 drill holes totalling approximately 207,000 m. In addition, four trenches were dug for a total of 60 m. At Cristinas, drilling was carried out by Placer from 1992 to 1997, consisting of 1,182 drill holes totalling approximately 155,000 m, and by Crystallex from 2003 to 2007, consisting of 90 holes totalling approximately 28,000 m. The Crystallex drill hole data was not available for RPA’s resource modelling work.

The Siembra Minera mineralization is open down dip in all zones and along strike to the northwest in Morrocoy and Cordova because of insufficient drilling. Current plans for exploration are based on brownfield expansion of the existing deposit. As the Project advances, GRE proposes to carry out approximately 75,000 m to 100,000 m of new drilling.

MINERAL RESOURCE ESTIMATES

A Mineral Resource estimate, dated December 31, 2017, was completed by RPA using the Surpac and Leapfrog Geo software packages. Wireframes for geology and mineralization were constructed in Leapfrog Geo based on geology sections, assay results, lithological information, and structural data. Assays were capped to various levels based on exploratory data analysis and then composited to three metre lengths. Wireframes were filled with blocks measuring 10 m by 10 m by 6 m (length, width, height). Block grades were estimated using dynamic anisotropy and inverse distance squared algorithms. Block estimates were validated

 

---

www.rpacan.com

using industry standard validation techniques. Classification of blocks was based on drill hole spacing distances and other criteria.

A summary of the Mineral Resources at the Project is provided in Table 1-7.

 

Notes:

1.	CIM (2014) definitions were followed for Mineral Resources.
2.	Mineral Resources are estimated at an NSR cut-off value of US$7.20 per tonne for oxide-saprolite material and US$5.00 per tonne for sulphide-saprolite and fresh rock material.
3.	Mineral Resources are constrained by a preliminary pit shell created using the Whittle software package.
4.	Mineral Resources are estimated using a long-term gold price of US$1,300 per ounce, and a copper price of US$3.00 per pound.
5.	Bulk density varies by material type.
6.	Mineral Resources that are not Mineral Reserves do not have demonstrated economic viability.
7.	Numbers may not add due to rounding.

RPA is not aware of any environmental, permitting, legal, title, taxation, socio-economic, marketing, political, or other relevant factors that could materially affect the Mineral Resource estimate.

MINING

The Siembra Minera Project is an open pit gold-copper mining project that will utilize 30 m³ hydraulic shovels and 236-tonne trucks as the primary mining equipment.

The resource pit optimization was developed by RPA based on the RPA Mineral Resource estimate (Table 1-7). Blocks classified as Measured, Indicated, and Inferred Mineral Resources were included in the resource pit optimization process for the Siembra Minera deposit. The resource pit is approximately 6,000 m long and 1,900 m wide with a maximum depth of approximately 700 m. The pit slope on the east wall follows the mineralization with slopes from 36° to 38°, while the west wall final pit has overall pit slopes ranging from 48° to 50°.

 

---

www.rpacan.com

Mine production is scheduled to be carried out at a maximum mining rate ranging from 330 ktpd to 380 ktpd of total material. Stripping ratios are expected to average 1.16 over the LoM plan. The production schedule was produced using Whittle software to guide the mining sequence, Vulcan to design phases, waste dumps and the final pit, and XPAC to schedule the phases following the processing requirements.

During the first ten years of the Project, 5.8 million tonnes per annum (Mtpa) of oxide saprolite that does not require grinding will be processed in the oxide saprolite plant. The flotation plant starts two years after the oxide plant. Feed to the flotation mill is scheduled for 58.0 Mtpa for years 3 to10, while softer high copper sulphide saprolite material is available. In year 11, one quarter of the flotation grinding mill (12.25 Mtpa) is converted to oxide to accommodate the harder low-copper sulphide saprolite and low-copper hard rock materials. The other 36.75 Mtpa of capacity in the grinding mill are used for the harder higher-copper material in the flotation. The oxide plant will start processing with a combination of saprolite and low copper hard rock using the leach tanks from the oxide saprolite plant and additional leach tanks required for processing. The hard rock and sulphide saprolite was divided into high copper and low copper using a 0.02% Cu threshold.

In order to supply the processing input required in the first 10 years of production, the total material mined must achieve up to 120 Mtpa from a combination of the mining phases. The mining rate will change depending on stockpile size, increasing total mining rate to 140 Mtpa in year 20.

Total resources potentially mineable by open pit are estimated at approximately 2.0 billion tonnes of mineralized material at a gold grade of 0.705 g/t and a copper grade of 0.1% with 2.3 billion tonnes of waste for a stripping ratio of 1.16 tonnes of waste per tonne of mineralized material.

All of the waste rock, except that used for TMF construction, will be disposed of in the WRD facilities located to the north, west, and south of the pit. It appears that a portion of the Siembra Minera pit could be backfilled with waste rock, however, further investigation into tailings disposal and pit backfill opportunities are recommended.

 

---

www.rpacan.com

MINERAL PROCESSING AND METALLURGICAL TESTING

The Siembra Minera Project consists of three rock types. Hard rock ore comprises approximately 87% of the material that will be processed. The remaining 13% of the mineralized material is saprolite with a split composed of approximately 43% oxide saprolite and 57% sulphide saprolite. Metallurgical test work was conducted on hard rock that contains higher and lower copper concentrations, and on blends that simulate the blends projected for the plant operation.

Based on the results of metallurgical testing, the conceptual processes selected for the combined project include a cyanide leach plant to process oxide saprolite and sulphide saprolite that contains low concentrations of copper to recover gold as doré from gravity concentration and cyanide leaching plus a flotation concentrator to process sulphide saprolite and hard rock that contain higher concentrations of copper. The flotation concentrator will recover copper and gold into a copper flotation concentrate and gold as doré utilizing gravity concentration and cyanide leaching of cleaner scavenger tailings.

The production schedule for this PEA is based on initially processing oxide saprolite through a 15,000 tpd cyanide leach plant. The crushing and screening plant feed is designed to process approximately 10% higher assuming that some of the material will be rejected due to oversize and/or rock material. Starting in year 7, the majority of the oxide saprolite is depleted and sulphide saprolite that contains low concentrations of copper will also be fed to the plant. In years 9 and 10, only low copper sulphide saprolite will be fed to the oxide plant.

In year 4, the flotation concentrator will be commissioned. The feed to the plant includes sulphide saprolite that contains higher concentration of copper and a combination of high and low copper hard rock material at a nominal rate of 140,000 tpd although the actual feed rate is higher in the early years due to the presence of sulphide saprolite which is easier to grind.

In year 11, the quantity of hard rock with suitable copper grades to produce acceptable concentrate grades in the flotation plant diminishes so the plant will be re-configured to process less material through the flotation plant and additional material through the oxide leach plant. The conceptual plan, at this early stage of the Project development, is to reduce the feed to the flotation concentrator to approximately 105,000 tpd and increase the tonnage to the oxide leach plant to 35,000 tpd. The low copper hard rock material will be ground in the existing milling circuit in the flotation plant and the leach plant will be expanded to accommodate the

 

---

www.rpacan.com

higher tonnage of material. The ball mill in the oxide leach plant, which is only sized to process saprolite, can be decommissioned or used to grind saprolite that is pumped from the open pit mine to the oxide leach plant.

ENVIRONMENT

Two separate, but parallel ESIA are being prepared for the Project. One ESIA is intended to meet Venezuelan regulatory requirements and the second one, international standards and guidelines. The Venezuelan ESIA is expected to be completed and submitted to the Ministry of People’s Power for Ecosocialism and Water (MINEA) in 2018; and the International ESIA will be completed soon thereafter.

Prior to submission of the ESIA, an Authorization to Occupy the Territory (AOT) must be obtained and a Term of Reference (TDR) approved. The AOT certifies that the proposed use of the land by the Project is compatible with the land use designation of the area and the TDR defines the scope and contents of the ESIA. Both AOT and TDR must be submitted to MINEA. GRE has submitted the application for an AOT, and the TDR for the Project will be submitted as soon as the AOT is approved. Upon the approval of the TDR, GRE will prepare and submit the ESIA to MINEA. An application for the Authorization to Affect Natural Resources (AANR), a permit for exploitation, will be submitted as soon as the Project ESIA is approved, which is expected to be in 2018.

In addition to the ESIAs, GRE is in the process of developing a series of environmental and social management plans and programs. Thousands of small-scale miners are actively working in the Project area and adequate management of small-scale mining is critical to the success of the Project. A conceptual plan for small-scale mining management has been developed by GRE to relocate these miners to the Oro concession area.

Based on the current Project design, reclamation activities will commence soon after construction begins, and will continue throughout the life of the Project. Closure activities will continue for three years after the end of the mine life in year 27. Some intermittent reclamation would also take place before year 23, when areas are no longer needed for mine operation activities. Total expenditures for reclamation and closure are currently estimated to be approximately US$150 million.

 

---

www.rpacan.com

CAPITAL COST ESTIMATE

A summary of capital costs is shown in Table 1-8.

 

OPERATING COST ESTIMATE

The Siembra Minera Project will process approximately 2,005 million tonnes of mineralized material over its planned 45-year mine life. The estimated average operating costs for the Project life are shown in Table 1-9.

 

---

www.rpacan.com

TABLE 1-9 ESTIMATED LOM OPERATING COSTS

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

Operating costs for this Project appear to be low, however, the diesel fuel price of $0.02/L, the electricity cost of $0.038/kWh ($38/MWh), and the low labour costs have a significant impact on the unit operating costs.

 

---

www.rpacan.com

2 INTRODUCTION

Roscoe Postle Associates Inc. (RPA) was retained by Gold Reserve Inc.(GRI), and its wholly owned subsidiary GR Engineering Barbados, Inc. (GRE) to prepare an independent Technical Report on the Siembra Minera Project (the Project), located in Bolivar State, Venezuela. The operating company Empresa Mixta Ecosocialista Siembra Minera, S.A. (Siembra Minera), which holds the rights to the Siembra Minera Project, is a mixed capital company with 55% being owned by a Venezuelan state entity [owned by the Bolivarian Republic of Venezuela through the Corporación Venezolana de Minería (CVM)] and 45% by GR Mining Barbados, Inc. (GRM), a wholly-owned subsidiary of GRI. GRE has been set up to perform engineering, procurement, construction, and operation of the Project.

The Project is a combination of the Brisas and Cristinas properties into a single project. The purpose of this report is to provide GRI and GRE with an initial assessment of the Siembra Minera Project including a resource estimate, conceptual mine plan, and a preliminary economic review. This Technical Report conforms to NI 43-101 Standards of Disclosure for Mineral Projects.

The Siembra Minera Project is a gold-copper deposit located in the Kilometer 88 mining district of Bolivar State in southeast Venezuela. Local owners and illegal miners have worked the property for many years. Shallow pitting and hydraulic methods were used to mine the upper saprolite zone, and coarse gold was recovered by gravity concentration and amalgamation with mercury. Most of the large-scale exploration work at Cristinas was performed by Placer Dome Inc. (Placer), which worked on the property from 1991 to 2001. At Brisas, GRI carried out the exploration program on the concession from 1992 to 2005. The most recent Technical Report for Cristinas is dated November 7, 2007, which is based on a feasibility study and includes historic mineral reserves. The most recent Technical Report for Brisas is dated March 31, 2008, which is also based on a feasibility study and includes historic mineral reserves.

RPA has relied on data derived from work completed by previous owners on the Cristinas concessions and by GRI on the Brisas concessions. The current resources for Cristinas were estimated by RPA based on the drill hole data supplied by Corporación Venezolana de Guayana (CVG) to GRI in 2002. The database had 1,174 drill holes and 108 trenches which were included in the Cristinas database. Hard copies of the assay data sheets were not

 

---

www.rpacan.com

available but GEOLOG data files from Placer were provided including assay data, geological descriptions, structural data, geotechnical data, and check sample data. The current resources for Brisas were estimated by RPA based on drill hole data supplied by GRI in Geovia GEMS format which formed the basis of the last Technical Report by Pincock Allen & Holt (PAH) in 2008.

This report is considered by RPA to meet the requirements of a Preliminary Economic Assessment (PEA) as defined in Canadian NI 43-101 regulations. The mine plan and economic analysis contained in this Technical Report are based, in part, on Inferred Mineral Resources, and are preliminary in nature. Inferred Mineral Resources are considered too geologically speculative to have mining and economic considerations applied to them that would enable them to be categorized as Mineral Reserves. There is no certainty that economic forecasts on which this PEA is based will be realized.

SOURCES OF INFORMATION

GRI provided access to their dataroom which included all the previous studies and mineral resource models. Discussions were held with personnel from GRI including:

• Mr. Doug Stewart, Project Manager
• Mr. Brad Yonaka, Chief Geologist

For this Technical Report, overall management was carried out by Richard J. Lambert, P.Eng., RPA Principal Mining Engineer. José Texidor Carlsson, P.Geo., RPA Senior Geologist, developed the mineral resource model under the supervision of Luke Evans, M.Sc., P.Eng., Principal Geologist. Hugo M. Miranda, P.C., RPA Principal Mine Engineer developed the pit optimization and production schedule. Kathleen A. Altman, Ph.D., P.E., RPA Principal Metallurgist, reviewed the metallurgical test work and process design. Grant A. Malensek, P. Geo., P. Eng., RPA Principal Valuation Engineer, was responsible for the Project economics. The site was visited by Mr. Miranda on September 19, 2017 and was previously visited by Mr. Lambert in February 2008.

Mr. Lambert is responsible for Sections 15, 16, 19 and 20 of this report and shares responsibility for Sections 1, 2, 3, 18, 21, 24, 25, and 26. Mr. Texidor is responsible for Sections 4 to 12, and 14 and shares responsibility for Sections 1, 2, and 23 to 26. Dr. Altman is responsible for Sections 13 and 17 and shares responsibility for Sections 1, 18, 20, 21, 24,

 

---

www.rpacan.com

25, and 26. Mr. Miranda is co-author for Section 16 and shares responsibility for Sections 1, 2, 3, 24, 25, and 26. Mr. Malensek is responsible for Sections 21 and 22 and shares responsibility for Sections 1, 2, 3, 24, 25, and 26. The documentation reviewed, as well as any other sources of information, are listed at the end of this report in Section 27 (References).

 

 

 

---

www.	rpacan.com

3 RELIANCE ON OTHER EXPERTS

This report has been prepared by RPA at the request of GR Engineering (Barbados).

The information, conclusions, opinions, and estimates contained herein are based on:

1.	information available to RPA at the time of preparation of this report,
2.	assumptions, conditions, and qualifications as set forth in this report, and
3.	data, reports, and opinions supplied by GRI and GRE and other third party sources.

For the purpose of this report, RPA has relied on ownership information provided by GRI. RPA has not researched property title or mineral rights for the Project and expresses no opinion as to the ownership status of the property.

RPA has relied on GRI for guidance on applicable taxes, royalties, and other government levies or interests, applicable to revenue or income.

Except for the purposes legislated under provincial securities laws, any use of this report by any third party is at that party’s sole risk.

 

---

www.rpacan.com

4 PROPERTY DESCRIPTION AND LOCATION

The Siembra Minera property is located in the Kilometre 88 mining district, Bolivar State, in southeast Venezuela at a Latitude of 6°10’ North and Longitude 61°28’ West. Highway 10, which is a major highway in Venezuela, is only 3.5 km from the property (Figure 4-1).

The Project site is located in the Guyana region, which covers approximately one-third of Venezuela’s national territory. The closest nearby large city is Ciudad Guayana, with approximately 1,050,000 inhabitants (2001), situated on the Orinoco River near its confluence with the Caroní River. Ciudad Guayana consists of the old town of San Félix to the east and the new town of Puerto Ordaz to the west. Puerto Ordaz is home to most of the major industrial facilities such as aluminum smelters and port facilities. Puerto Ordaz has major port facilities accessible to ocean-going vessels from the Atlantic Ocean via the Orinoco River, a distance of approximately 200 km. There is regularly scheduled commercial airline service to Puerto Ordaz from Caracas.

Ciudad Guayana is the centre of major industrial developments in the area, including iron and steel mills, aluminum smelters, iron and bauxite mining, and forestry. These industries are supported by major dams and hydroelectric generating plants on the Caroní River, providing 12,900 MW of electricity.

LAND TENURE

The Project survey control is based on the Universal Transverse Mercator (UTM) coordinate system. It is based on the Zone 20 North projection, using the World Geodetic System 1984 (WGS’84) datum. Surco, S.A. (Surco), a local survey firm based in El Callao, Venezuela, established permanent survey reference points within the Project area. The base for all surveys was Global Positioning System (GPS) survey, defined and checked by the survey company with a traverse from a nearby GPS station (Cristinas) with satisfactory accuracy. Surco surveyed the drill holes for both Placer and GRI.

The Siembra Minera Economic Zone (Project boundary) occupies a rectangular area of approximately 18,951 ha. The dimensions of the property are 20.5 km (north-south) by 11.2 km (east-west). The Economic Zone will contain the open pit mine, all Project infrastructure,

 

---

www.rpacan.com

and waste rock dumps (WRD) and tailings management facilities (TMF). The land includes the prior concessions of Cristinas 4, Cristinas 5, Cristinas 6, Cristinas 7, Oro 1, Albino 1, Brisas del Cuyuni, Carabobo, Morauana, Barbara, Zuleima, El Pauji, Esperanza, and portions of Guarimba, Mireya, Virgen de Lourdes, Lucia, and a few smaller concessions. The Economic Zone is shown in Figure 4-2.

The Economic Zone is designated by Presidential Decree No. 30 dated October 31, 2016 and authorized by Nicolás Maduro Moros. The decree delimits the geographic area in which Siembra Minera shall perform activities of exploration and exploitation of gold mines and deposits, including their production. The coordinates of the Economic Zone are presented In Table 4-1.

TABLE 4-1 UTM COORDINATES OF ECONOMIC ZONE

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

RPA is not aware of any environmental liabilities on the property other than existing mercury levels from artisanal miners as discussed in Section 20. RPA is not aware of any other significant factors and risks that may affect access, title, or the right or ability to perform the proposed work program on the property.

 

---

www.rpacan.com

 

4-3

---

 

---

www.rpacan.com

5 ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY

ACCESSIBILITY

El Dorado (population est. 5,000, 2011), which is 88 km north of the Project site on Highway 10, is historically the nearest population centre of any size. Over the past ten years, the town of Las Claritas (population est. 6,000, in 2011, previous est. 2,500, in 2001) adjacent to the Project has surpassed El Dorado with the influx of artisanal miners. Las Claritas is a small town located on Highway 10.

Highway 10 provides paved access from Ciudad Guayana and Puerto Ordaz, which is 373 km northwest of the property, to within 3.5 km of the Project site. Unpaved roads provide the remaining 3.5 km of access. Upgrading the unpaved roads is part of the infrastructure improvement plans for the Project area which will include three main Project access routes, one from the north, one from the east near Las Claritas, and a third to the south providing direct access to the process plant from Highway 10.

CLIMATE

The climate is tropical with January through March being drier months and June through July being wetter months. Humidity is high (monthly average 80-87%) and annual precipitation is over 3,000 mm. Temperatures are fairly uniform with average highs around 35°C and average lows around 23°C. Daily temperatures range from 21°C to 38°C. Prevailing winds are from the west – southwest, with a speed mostly from 0.5-2.1 m/sec. There are extensive plans for surface and ground water control so that mining can be conducted year-round.

LOCAL RESOURCES

Plans are to construct a dual use camp facility at a location adjacent to the process area. This camp will initially house construction workers and at the conclusion of construction will be converted to a permanent facility for a portion of the operating personnel. The construction camp configuration will have a nominal capacity of 2,400 men based on an occupancy rate of two men per room. This may be increased during peak periods if required.

 

---

www.rpacan.com

When the construction has been completed the camp will be re-configured to provide a single room for each man and permanent recreation facilities for the operating staff. The permanent operations camp facility will have a capacity of 1,200 staff. It is assumed that the rest of the operating staff (~300 personnel) will come from the local communities.

Busing will be considered if required from El Dorado and other communities.

INFRASTRUCTURE

To support the mining and milling operations at the Siembra Minera Project, a number of ancillary facilities will be required. These include a mine equipment maintenance shop, warehouse, reagent storage building, laboratory, and administration offices. A construction camp will be prepared and will be converted to an operation camp. The operational man camp size is based on the assumption that approximately half the work force is away on scheduled time off due to crew rotations. Previously, there was a small camp with several cinder block buildings that could house up to 100 people at Brisas and a larger camp at Cristinas that was constructed by Placer. The camp at Brisas was used to support the exploration programs and the camp at Cristinas was used to support the initial construction efforts. Both camps have been destroyed and a new larger camp will be constructed near the plant site.

Three unpaved roads are used to access the Project from Highway 10. Plans are to improve these to provide access to the mining area and the process area. A network of service roads will be constructed to allow access to the camp facility, tailings dam, sedimentation ponds, explosives magazine, and other remote installations. Major deliveries will use either the north access road or the south access road and will avoid the Las Claritas village.

A water supply and distribution system will be constructed, using the pit dewatering wells as a source of fresh water. The mill area, mining area, and the campsite will each be provided with a sewage collection and treatment system.

The power authority, Corporación Eléctrica Nacional S.A. (Corpoelec) is the fully integrated state power corporation of Venezuela. It was created in 2007 by merging ten state-owned and six private-owned power companies. Corpoelec constructed a power line south from Ciudad Guayana into Brazil. The authority has also constructed a substation at Las Claritas located approximately 3.5 km from the Project, which has sufficient power to supply the Siembra Minera Project.

 

---

www.rpacan.com

PHYSIOGRAPHY

The Project area is located at the foot of the Sierra de Lema high plateau; and, the topography is moderately homogenous, dominated by plains with some rolling hills. Terrain in the mine area is relatively flat with elevations ranging from 120 MASL to 220 MASL with higher elevations near the east and southeast margins of the Project area. Near the plant site and tailings disposal areas, the terrain goes from being relatively flat to fairly steep. The tailings disposal site design has used this as an advantage by constructing the dam in the flat area and using the hillside as the back of the tailings disposal facility. Elevations in the plant and tailings disposal area range from 130 MASL to 200 MASL. The plant site will be at 190 m above sea level.

Most of the area is covered by moderately dense sub-Amazon rainforest. Trees range from 25 m to 35 m in height. Low-lying areas tend to be wet and swampy. There are many pits left by artisanal miners that are filled with water. They will require pumping prior to mining.

 

---

www.rpacan.com

6 HISTORY

EXPLORATION AND DEVELOPMENT HISTORY

General Fernandez Amparan first discovered gold in the Siembra Minera region in 1920. Gold mining at the site was initiated in the 1930s and continued sporadically on a minor scale until the early 1980s when a gold rush occurred. Approximately 5,000 to 7,000 small miners worked alluvial and saprolite-hosted gold deposits using hydraulic mining techniques. Many square kilometres of jungle have been stripped of soil and saprolite. This material was processed in sluices and small hammer mills. The amount of gold recovered is unknown and much of the area of the concessions is now covered with tailings. The name Kilometre 88 for the district came from the area being located near kilometre 88 marker of the road linking El Dorado with the Brazilian border (Pan American Highway or Highway 10).

Placer conducted essentially all of the modern exploration on Cristinas during their tenure on the property from 1991 to 2001. Placer completed line cutting, mapping, rock and soil sampling, geophysics, and drilling of 1,174 drill holes for a total of 158,738 m of drilling. After extensive exploration, Placer announced commencement of construction of the Las Cristinas mine on August 2, 1997. The inauguration took place at the site with officials of Placer, CVG, and representatives of the Venezuelan Government present. On January 20, 1998, Placer announced that its operating company in Venezuela, Minera Las Cristinas C.A., had decided to suspend construction. Construction resumed in May 1999 but was again suspended on July 15, 1999 due to uncertainty with respect to gold prices and title. Up until that time, Placer had reportedly spent US$168 million on the Project. CVG took possession of the Cristinas property in 2001 and in 2002 signed a mine operating agreement (MOA) with Crystallex International Corporation (Crystallex) whereby Crystallex was required to explore, mine, and produce gold at Las Cristinas.

Crystallex undertook drilling to confirm results of the previous operator prior to their first resource estimate. Crystallex drilled 12 holes totalling 2,199 m in 2003 to confirm the tenor of mineralization presented in the pre-existing database and also assayed check samples. For additional confirmation, Crystallex re-assayed 262 pre-existing pulps, 200 pre-existing coarse rejects, and 342 pre-existing quarter-core samples. During the course of the drill data verification and the resource expansion drilling in 2005, it was noted (MDA 2005) that some biases existed between Crystallex and Placer data, the latter of which represent by far the bulk

 

---

www.rpacan.com

of the exploration data. A heterogeneity study was undertaken to better understand the grade biases noted, to define more appropriate sub-sampling procedures and protocol, and to maximize the efficiency of the upcoming grade-control program during mining operations. A report by Francis Pitard (2005) suggested that the grade bias of Crystallex grades being lower than Placer grades was likely due to the difference in size of the core samples. Pitard further pointed out that the samples taken by Placer also could be understating the global grade of the Cristinas deposit. Crystallex completed a 46-hole drill program in February 2007. The audits concluded that Placer and Crystallex procedures met or exceeded industry standards at the time, and assay laboratories provided reliable and acceptable results.

In February 2011, the MOA was terminated by CVG.

The Brisas concession was acquired by GRI in August 1992 with the acquisition of Compañia Aurifera Brisas del Cuyuni C.A. Prior to 1992, no known drill holes existed on the Brisas site. Initial work by GRI included surface mapping, regional geophysical surveys, and geochemical sampling. Several anomalies were identified on the property and drilling and assaying began in 1993. A large deposit with stratabound gold-copper mineralization was discovered in both alluvial and hard rock material early in the drilling program. Additional work followed with petrology, mineral studies, density tests, metallurgical sample collection, and laboratory test work.

Initial exploration drilling by GRI commenced in 1993 utilizing both auger and core drilling methods. Most of the exploration and development drilling took place in 1996 and 1997. From 1996 on, all exploration drilling has been completed utilizing diamond drill core rigs. Additional exploration drilling was completed in 1999, 2003, 2004, and 2005. As of 2005, 802 exploration holes had been drilled of which 731 are diamond core holes. This represents 186,094 m of core drilling, and 189,985 total m of exploration drilling, core and auger. All split core was stored on site until 2008, but has since been ransacked and displaced.

Since 2005, an additional 76 holes have been drilled on the Brisas concessions for geotechnical and other studies. These holes have not been included in any resource modelling because they were not drilled for exploration purposes.

Independent verification by Behre Dolbear & Company. Inc. (Behre Dolbear) of drilling, assaying, and data collection procedures was undertaken in 1997 and verification of the

 

---

www.rpacan.com

computer database, mine modelling procedures, and reserve estimate was completed in 1998. The audits concluded that GRI procedures met or exceeded industry standards, and assay laboratories provided reliable and acceptable results.

In August 2016, GRI signed a mixed company agreement with Venezuela for the formation of a jointly owned company and, in October 2016, established Siembra Minera, a mixed capital company with 55% owned by a Venezuelan state entity and 45% by GRM, a wholly-owned subsidiary of GRI.

HISTORIC RESOURCE ESTIMATES AND FEASIBILITY STUDIES

CRISTINAS

Resource and reserve estimates for the Cristinas deposit were completed by Mine Development Associates (MDA) on April 30, 2003. These results were filed on SEDAR as the Technical Report titled “Resources and Reserves, Las Cristinas Gold and Copper Deposits, Bolivar State, Venezuela” prepared by MDA. The measured and indicated resource was estimated at 439 million tonnes with an average gold grade of 1.09 g/t for a total of 15 million contained ounces based on a gold cut-off grade of 0.5 g/t. Proven and probable mineral reserves were estimated at 224 million tonnes with an average gold grade of 1.33 g/t containing 9.54 million ounces (MDA, 2003).

Subsequent to the filing of the 2003 Technical Report by MDA, there have been other resource and reserve estimates released by Crystallex. A feasibility study for Las Cristinas was completed in 2004 and a Development Plan, in 2005 by SNC-Lavalin. An updated Technical Report was filed on SEDAR in August 2005 based on the new Development Plan. The 2005 Technical Report shows proven and probable reserves at Las Cristinas of 294 million tonnes grading 1.32 g/t Au for a total of 12.5 million contained ounces of gold (SNC-Lavalin, 2005).

An updated resource and reserve estimate and a Technical Report were completed for the project by MDA in conjunction with SNC-Lavalin on November 7, 2007. These results were filed on SEDAR as the Technical Report titled “Technical Report Update on the Las Cristinas Project, Bolivar State, Venezuela” prepared by MDA. The measured and indicated resource was estimated at 629 million tonnes with an average gold grade of 1.03 g/t for a total of 21 million contained ounces based on a gold cut-off grade of 0.5 g/t. Proven and probable mineral reserves were estimated at 464 million tonnes with an average gold grade of 1.13 g/t containing 16.86 million ounces of gold (MDA, 2007).

 

---

www.rpacan.com

The resource and reserve estimates quoted above are considered to be historic in nature and should not be relied upon, however, these are relevant as they indicate the presence of mineralization on the Project.

BRISAS

J.E. MinCorp, a division of Jacobs Engineering Group, Inc. completed a pre-feasibility study on the Brisas Project in February 1998. In addition, a supplement to the pre-feasibility study was completed in August 1998, addressing the merits of the Cominco Engineering Services Ltd. (CESL) hydro-metallurgical process. The CESL process was a method of treating copper concentrates on site by pressure oxidation, acid leaching with solvent extraction/electrowinning recovery of copper in the form of copper cathode, and gold recovery by a cyanide leach of the solids. Work completed since 1998 and directed at project optimization includes updating the mine computer model and ultimate pit designs, mine planning and optimization of cut-off grades, and updated slope stability design criteria. In addition, work was completed on mill tailings characterization and analysis of physical properties, cyanide destruction test, and settling and thickening tests for plant design criteria.

In 2003, Behre Dolbear completed Mineral Resource and Mineral Reserve estimates for the Brisas deposit and documented it in a Technical Report filed on SEDAR. The estimates were based on two scenarios for treating the copper concentrates, a conventional smelter and refining case and the use of CESL hydro-metallurgical process (Behre Dolbear, 2003). Using a gold price of $325 per ounce and a copper price of $0.85 per pound, proven and probable reserves were estimated to be 257 million tonnes grading 0.805 g/t Au and 0.135% Cu containing 6.64 million ounces of gold and 764 million pounds of copper for the smelter case and 328 million tonnes grading 0.708 g/t Au and 0.150% Cu containing 7.48 million ounces of gold and 1.08 billion pounds of copper for the CESL case.

GRI commenced work for a bankable feasibility study in the last quarter of 2003. Several major engineering and consulting companies were selected to complete the work necessary for the feasibility study. They were Aker Kvaerner, an engineering and construction company specializing in mining and mineral processing; Vector Colorado, a tailings dam design, geotechnical and hydrology specialist; and PAH for the mineral resource and reserve estimate, pit design, mine planning, and mine cost estimation. This feasibility study was completed in January 2005. In addition, AATA International and Ingenieria Caura, S.A. was selected to complete an Environmental and Social Impact Assessment (ESIA) Study in compliance with

 

---

www.rpacan.com

the World Bank and International Finance Corporation (IFC) Standards for meeting Equator Principles criteria. The feasibility study operating plan assumes a large open pit mine containing proven and probable reserves of approximately 9.2 million ounces of gold and 1.2 billion pounds of copper in 414 million tonnes of ore grading 0.69 g/t Au and 0.13% Cu (PAH, 2005).

The combination of the 2005 feasibility study and subsequent studies provided the basis of the 2008 Technical Report. Several optimization studies were conducted to determine the most economic process plant option and production rate. Total proven and probable mineral reserves for the Brisas Project in 2008 were estimated at 482.7 million tonnes grading 0.66 g/t Au and 0.13% Cu. A total of 1.08 billion tonnes of waste was estimated in the pit resulting in a stripping ratio (waste:ore) of 2.24:1.0 (PAH, 2008).

All Mineral Reserve estimates quoted above are considered to be historic in nature and should not be relied upon, however, these are relevant as they indicate the presence of mineralization on the Project.

All previous Mineral Resource estimates for the Brisas Project are superseded by the current Mineral Resource estimate in Section 14 of this report.

PAST PRODUCTION

There are no records of previous gold production from artisanal miners.

 

---

www.rpacan.com

7 GEOLOGICAL SETTING AND MINERALIZATION

REGIONAL GEOLOGY

The Siembra Minera Project is located within the Guyana Shield in northern South America. The shield covers easternmost Colombia, southeastern Venezuela, Guyana, Suriname, French Guiana, and northeastern Brazil. The Venezuelan portion of the shield is subdivided into five geological provinces with different petrological, structural, and metallogenic characteristics. The provinces are, from oldest to youngest, Imataca, Pastora, Cuchivero, Roraima, and Parguaza. Only the Imataca, Pastora, and Roraima provinces are found in the vicinity of the Siembra Minera deposit.

Rocks of the Imataca Province constitute the oldest terrain in the Venezuelan Guyana Shield and include quartzo-feldspathic gneiss, felsic, and mafic granulites, and iron formation. This province is located along the Orinoco River in the northern portion of the Guyana Shield. Rocks in the terrain are tightly folded, highly metamorphosed, and have ages ranging from 3,700 Ma to 2,150 Ma. The oldest age represents the protolith, whereas the younger age represents the Trans-Amazonian orogeny of Lower Proterozoic age. The Imataca Province is known for iron deposits hosted by banded iron formations.

The Pastora Province is separated from the Imataca terrain by the Guri fault on its northern edge and extends to the Kilometre 88 gold district in the south. This province is characterized by several penecontemporaneous tholeiitic and calc-alkaline volcano-sedimentary sequences. Rock types that have been described and are not necessarily present in all sequences include pillow basalt, andesite, dacite, rhyolite, tuffaceous and pyroclastic sediments, greywacke, pelite, tuff, and chemical sedimentary rocks. Rocks of the province were metamorphosed to greenschist facies and intruded at various levels by granitic rocks of the Supamo Complex (2,230 Ma to 2,050 Ma). This petrologic assemblage constitutes the granite-greenstone belts of Lower Proterozoic age, which extends into Guyana, Suriname, French Guiana, and Brazil. The Trans-Amazonian orogeny (2,150 Ma to 1,960 Ma) was a period of deformation, metamorphism, magnetism, and enrichment of previously deposited gold-bearing volcano-sedimentary rocks in the Venezuelan part of the Guyana Shield as well and in the other mentioned countries. Rocks of this province have been intruded by Lower Proterozoic (1850

 

---

www.rpacan.com

Ma to 1650 Ma) and Mesozoic (210 Ma to 200 Ma) diabase dikes, sills, and gabbroic bodies related to crustal extension.

The Roraima Province of Middle Proterozoic age (1,600 Ma) is exposed to the south of the Kilometre 88 district. This province includes sedimentary rocks of continental origin that were laid unconformably on top of the granite-greenstone terrain. These rocks are not metamorphosed, have horizontal to low angle dips, and are intruded by Mesozoic diabase dikes and sills.

LOCAL AND PROPERTY GEOLOGY

The greenstone belt present in the Kilometre 88 district consists of four formations, listed below oldest to youngest:

1) Lower Carichapo Group meta-lavas, meta-tuffs, amphibolites, and ferruginous quartzites.

2) Lower Proterozoic greenstone basalts, andesites, tuffaceous rocks, pyroclastic breccias, and metagraywackes. These rocks are lithologically similar to the Caballape Formation defined in the Botanamo district to the north east, but geographically isolated. For convenience, they are referred to as Lower Caballape in this report.

3)	Granites and granodiorites of the Supamo Complex.
4)	Diabasic and gabbroic dikes and sills of Lower Proterozoic and Mesozoic ages.

The position and coverage of the above units have been established, at least on a regional scale, through aerial photos. Ground reconnaissance by government missions and more recently by private entities has either confirmed or mapped modifications to the aerial interpretations. The present geologic map is a composite of the above work. Rocks of the Carichapo Formation surround the concessions to the southwest, southeast, east, and north. They generally correspond to areas of higher topographic expression and are not commonly host to significant gold deposits. Greenschist volcanic and volcano-sedimentary rocks of calc-alkaline composition (called Lower Caballape Formation in this report) constitute the major units present in the areas of gold deposits, including the Brisas and Cristinas properties. The older unit covering the concessions consists primarily of intermediate tuffaceous rocks, and the younger unit to the west consists of intermediate to felsic tuffs, lavas, and volcano-sedimentary rocks. This sequence of rock units corresponds to areas of low, flat topography, forming hills only where the rock mass is more silicified. Relatively unfoliated intrusions of Supamo Complex granites are restricted to the south, east, and northeast, where they are

 

---

www.rpacan.com

topographically indistinct from the greenschist volcanics. All of the above units are intruded by Lower Proterozoic and Mesozoic diabases and gabbroic bodies, both as large mappable features, and as thin dikes and sills occurring in the volcanic units.

The Siembra Minera deposit lies within a portion of the lower Caballape Formation volcanic and volcanic-related sedimentary rocks. The units present are (1) andesitic to rhyolitic tuffaceous volcanic beds, (2) related sedimentary beds, and (3) a tonalitic intrusive body (Figure 7-1). All rocks have been tilted and subjected to lower greenschist facies metamorphism. It is thought, based on information from nearby properties, that the Siembra Minera Project occupies one limb of a large regional fold. Limited direction-indicating structures show the strata to be top-up. In the main mineralized trend, moderate to strong foliation is oriented N10°E and dipping 30° to 55° northwest. This foliation appears to be parallel to the original bedding and tends to be strongest in the finer-grained rocks. A much weaker foliation orientation appears in outcrop exposures, striking north-northwest and dipping to the southwest.

Dikes and quartz veins cut the lower Caballape Formation. The strata and intrusive rocks are cut by N30°W striking mafic dikes emplaced at regular intervals (200 m to 600 m), some of which have displacement in the order of tens of metres. These dikes are thought to be related to the Mesozoic diabase intrusions present throughout the district. Quartz veins populate the concession and have been noted both in outcrop and in drill intersection. The most common are sets of thick, boudinaged, and en echelon vein structures that follow foliation/bedding orientation. They are thought to relate in part to movement of quartz during metamorphism. Other quartz veins exist in various orientations that cannot be definitively linked to the structural elements described above.

One of the largest and best-defined stock reaches surface, in the saprolite, in a northeast-trending zone in the Potaso area on the south edge of the Cristinas deposit. The diorite, located north of the Potaso area, is asymmetric in a north-south section: it has a sub-vertical northwest face while its roof is shallowly inclined, dipping south at an angle of approximately 30° beneath the northern edge of the Brisas de Cuyuni deposit. This diorite stock occupies the gap in economic mineralization between the Cristinas and Brisas de Cuyuni deposits. The second diorite stock is located in the northern part of the Cristinas concessions, where it occupies the gap in mineralization between the Mesones and Morrocoy areas.

 

---

7-4

---

www.rpacan.com

ROCK UNITS

There are two general categories of rock units: weathered and unweathered rock. Weathered rock is further defined by degree of oxidation and mineral replacement due to weathering processes. Unweathered rock is further defined by lithology into various subdivisions of volcanic extrusive or intrusive units.

WEATHERED ROCK AND SAPROLITE

Oxidized Saprolite. A red-brown to yellow saprolite occurs in almost all parts of the concession from the surface to an average depth of 24 m. It is absent in the few areas where hard rock material outcrops. It is composed of clays, quartz, and hard ferruginous material in which all sulphide minerals have been oxidized and most other rock-forming minerals have been broken down to clay minerals and quartz.

Sulphide Saprolite. Sulphide saprolite, varying in thickness from less than one metre to 80 m, occurs immediately below oxidized saprolite. The water table constitutes the contact between the two and is generally sharp. It is noted on the geologic logs as “BOS” (base of oxidized saprolite). Sulphide saprolite is predominantly clay with both primary and secondary sulphides, the original rock having been broken down beyond recognition. Fragments of hard tuffaceous rock can occur. The initial occurrence of hard rock fragments in this unit (or in oxidized saprolite) is denoted on the geologic logs by the acronym “BAS” (base of 100% clay material). This boundary can exist in either sulphide or oxidized saprolite. The sulphide saprolite is well developed in the mineralized zone of the concession, but can be quite thin or absent in areas distal to mineralization.

Weathered Rock. Weathered rock is a label for any hard rock existing in a state of intense weathering, but not sufficiently broken down into clay to qualify as a saprolite. In general it falls between two contacts noted on geologic logs as “BZM” (base of mixed clay/hard rock material) and ‘BDM’ (base of weathering). In practice it is logged as the original rock type or as schist in the event that the original texture cannot be distinguished. Below the BDM, rock exists in a state of weathering in which the only chemical change is the leaching of calcite. The base of this layer is denoted as “BDL” (base of leaching), and below the rock is considered completely fresh.

 

---

www.rpacan.com

UNWEATHERED ROCK

Schist Units. The classification of schist is used when the original tuffaceous texture of the rock units has been erased by metamorphic processes. Schistosity is developed parallel to bedding, so schist units generally, but not always, follow dip of the tuffaceous units. Two types of schist have been defined: chlorite-sericite-biotite schist and quartz-sericite-pyrite schist.

Volcanic Units. The original unweathered rock types are calc-alkaline volcanic tuffs, generally of andesitic to dacitic composition. Occurrences of tuffaceous units reworked by sedimentary processes have been noted, but not to any great extent. Nomenclature of tuffaceous units has been established through observation of core, petrographic analysis, and geochemical data. Bedding and, to a lesser extent, graded bedding are commonly recognized. In general, feldspar crystal abundances are counted only with crystals exceeding one millimeter in diameter, and the field term of a “lapilli” is a pyroclast exceeding two millimeters in diameter. The ternary diagram provided in Figure 7-2 illustrates the composition of the various volcanic rock types recognized on the concessions.

a)	Vitric Tuff. Vitric tuff (TV) is a fine-grained, crystal-poor tuffaceous volcanic rock usually black in colour where not highly sericitized. It consists predominantly of glassy material, now devitrified, from the fallout of ash-sized particles. By definition it contains less than 10% feldspar crystals and less than 10% lithic fragments. It varies from a finely-banded volcanic sediment, to more massive mud-flow type deposit, which may contain lapilli pyroclasts, to a fine-grained massive texture. It is fully gradational into TVC-M and TL units (defined later).
b)	Crystal-Vitric Tuff. Crystal-vitric tuff (TVC-M) is defined as a tuffaceous unit having 10% to 40% feldspar crystals, and less than 10% lithic fragments. Locally the crystal content can drop as low as 10% but averaged over an entire depositional unit must exceed 10%. The upper boundary of 40% crystals is arbitrarily set, local fluctuations being ignored. When lapilli are observed and amount to more than a few percent of the rock mass, the unit is described as a lapilli-bearing TVC-M or TV.
C)	Crowded-Crystal-Vitric Tuff. Crowded-crystal-vitric tuff (TVC-C) is defined as a tuffaceous unit having greater than 40% feldspar crystals and less than 10% lithic fragments. It commonly contains significant mafic minerals (e.g., amphibole altered to biotite). If more than a few percent lithic fragments are observed, the unit is described as a lapilli-bearing TVC-C. Crowded-crystal-vitric tuff commonly resembles andesite porphyry, but numerous small lithic lapilli and grain size variations refute this possibility.

 

---

 

7-7

---

www.rpacan.com

d)

Lithic Tuff. Lithic tuff (TL) is defined as a tuffaceous unit containing greater than 10% lapilli-sized fragments. This definition is used without regard to presence or absence of feldspar crystals in the matrix, as field rock descriptions do not allow for further textural distinction. The fragments in some cases appear to be pieces of tuffaceous rock, presumably torn from its location by later volcanic activity. Pumice fragments have also been noted. It has been found to be important as a marker horizon, as it has an unmistakable texture and for the most part is observed in thin but easily definable units.

Intrusive Units. There are three mineralogically and texturally distinct occurrences of intrusive units, which vary from basaltic to dioritic in composition, all of which, are younger than the tuffaceous units described above.

a)	Mafic Dikes. This fine-grained, probably hypabyssal rock has a prominent ‘spinifex’ texture defined not by olivine, but rather by feldspar grains. They are unaltered, unfoliated, and magnetic. There are six such dikes on the concession, striking generally N40W, spaced 200- to 600-meters apart. They range from less than one metre to over five metres in width. Cross-cutting relationships indicate that they are the youngest rocks on the Concessions.
b)	Intermediate Coarse-Grained Intrusive. A coarse-grained tonalitic intrusive has been identified in only one area in the eastern part of the Concessions. It appears to be amorphous in shape and drilling has not encountered a lower contact. It is a coarse- grained, equigranular rock in large part unfoliated, but cut by discrete zones of strong deformation, both with and without sulphides and alteration. Zones of fracture- controlled chalcopyrite are also present, though the body does not exhibit economic Au or Cu mineralization. The only contacts observed to date are with TVC-C and are difficult to pinpoint as the two units can appear similar in hand sample. In one drill hole, it is cut by a mafic dike. The equigranular texture, high quartz content, and grain size are diagnostic. TVC-C, with which, it is sometimes confused, tends to have much greater variation in crystal content.
c)	Intermediate Aphanitic Sill/Dike. Intermediate hypabyssal intrusives occur as sill- like bodies less than one metre thick. These intrusives are usually aphanitic and are weakly foliated. They are useful as marker horizons within the volcaniclastic pile.

 

---

www.rpacan.com

STRATIGRAPHY

A stratigraphic column for the concession has been defined from the lithologic interpretation of the drill holes and is presented below in outline form (oldest to youngest):

a)	The lowest grouping is a sequence of crystal and crowded-crystal tuffaceous units that have a uniform appearance with very gradual changes in crystal percentages. The base of this sequence has not been reached by drilling.
b)	A thick crystal-vitric tuff and underlying vitric tuff that appears in the Cristinas concessions and the northern part of the Brisas concession (north of 682,500 N).
	South of this line the unit either pinches out, or drilling has been insufficient at depth to properly define it.
c)	A 150 m to 200 m thick sequence consists of rapidly alternating TL, TV, and TVC-M units. A prominent band of TL defines the base. Within this group only the TL bands and one TV bed are found to be laterally continuous, though even they are highly variable in thickness and extent. The bulk of economic gold mineralization occurs within this sequence. A sill of intermediate composition exists near the base of this sequence and is traceable throughout the concessions. The entire sequence thins toward the south, narrowing to less than 100 m at 681,500 N.
d)	A TV unit greater than 200 m thick appears throughout the concessions and contains minor TVC-M and TL bands. Much of this unit has a very even texture, and the contact with the underlying unit is readily apparent in most drill holes.
e)	A poorly defined sequence of TL, TV, TVC-M, and TVC-C units overlies (D), but is well outside, the mineralized zone and only encountered in a few condemnation drill holes to the west. This area has a strongly developed foliation, to the point where many units have been lumped together as “schist.”
f)	A diorite/tonalite intrusive feature exists on the eastern edge of the concessions that appears to postdate emplacement of the tuffaceous units, as it cross-cuts the stratigraphy, however, information about the contact between this body and the tuffaceous units is limited. No strong mineralization has been discovered in or at the margins of this body.

Regional mapping by the Venezuelan Geological Survey shows the Cristinas Project lying within the Caballape Formation of the Botanamo Group. The Caballape Formation is described as consisting largely of graded wackes and other sedimentary facies with minor andesitic to rhyodacitic volcanic intercalations. This description contrasts with the dominantly mafic to intermediate composition volcanic nature of the sequence that hosts the mineralization at Cristinas. The host sequence at Cristinas is now considered to constitute part of the Carichapo Group of the Pastora Supergroup (Table 7-1).

 

---

 

 

ALTERATION

Alteration of the original rock-forming minerals, such as amphibole and feldspar, and addition of elements such as boron and sulphur, is a result of hydrothermal, metamorphic, and weathering processes. The overprinting of these three processes has created a number of gradational alteration assemblages, which include varying amounts of quartz, secondary biotite, chlorite, sericite, calcite, epidote, metallic sulphides, tourmaline, magnetite, and minor fuchsite and anhydrite.

 

---

www.rpacan.com

Hydrothermal alteration is most intense within the Blue Whale body, and in other isolated pockets of similar appearance scattered throughout the main mineralized trend. The alteration type of the breccia approaches a greisen, with components of phyllic alteration in the schist. In many cases within the breccia pipe, fragments have been completely replaced by tourmaline, and associated zones of quartz may be a result of tourmalinitization of feldspars. Petrographic analysis shows two separate phases of growth in some tourmaline crystals. Massive occurrences of sulphides typically show an earlier phase of pyrite formation with subsequent fracturing and infilling of fractures by chalcopyrite.

Weaker propylitic alteration is present in tuffaceous units surrounding the Blue Whale body as strong calcite+epidote+pyrite and calcite+chlorite+pyrite+epidote+chalcopyrite assemblages. Typically, in lenses of high Cu/low Au mineralization, the alteration package is more potassic (high secondary biotite+chlorite±sericite). Many veins with these alteration assemblages are highly deformed, indicating emplacement prior to metamorphism.

Metamorphic alteration occurs throughout the concession and is thought to be the result of regional burial. Petrographic analysis identifies both biotite grade and chlorite grade metamorphic facies, occurring in the lower mesozone and upper epizone, respectively. This corresponds to a temperature range of 300°C to 500°C, and hydrostatic pressures. The gold+pyrite±Cu disseminated lenses appear to be associated with fluids present during this metamorphic event. The primary orientation of schistosity is thought to be parallel to bedding, with a weakly developed secondary schistosity at about 10° to bedding. Some chlorite and epidote formation may be attributed to subsequent retrograde metamorphism. Overprinting this initial metamorphism is an alteration assemblage possibly related to a tensional event that resulted in the development of barren calcite±quartz veins.

Weathering has resulted in the breakdown of the above mineral assemblages according to their compositions, ultimately resulting in the formation of smectite, illite, and kaolinite. Pyrite is retained in the unoxidized material, though is typically very fine grained and sub- to euhedral, suggesting secondary formation. Chalcocite is present in areas of high copper. Above the water table iron oxides have formed after sulphide minerals, releasing free gold. The assemblage most resistant to this process is the Blue Whale breccia, due to the high silica and tourmaline content.

 

---

www.rpacan.com

MINERALIZATION AT BRISAS

There are four distinct types of gold and copper mineralization present at Brisas, defined by geometry, associated minerals, and the gold-copper ratio. These zones are the Blue Whale body, disseminated gold + pyrite ± copper, disseminated high copper, and shear-hosted gold. Only the first three types are encountered within the proposed pit geometry. A more detailed description of the mineralization follows.

THE BLUE WHALE BODY

The Blue Whale mineralized body is a discrete, sharply bounded, flattened, cigar-shaped feature that trends more or less parallel to the local schistosity and plunges approximately 35° southwest. It outcrops in the Pozo Azul pit in the northeast portion of the Brisas concessions, and is intersected by 45 drill holes. It is 20 m in diameter at its widest point, and tapers off at depth. It is volumetrically a small fraction of the economically mineralized ground at the Siembra Minera Project, but it possesses the highest gold and copper grades.

Mineralogically, the Blue Whale is a sericite-tourmaline-pyrite-chalcopyrite-quartz schist, with a smaller volume of quartz-tourmaline-sulphide breccia. The schist is fine-grained and exhibits an almost complete alteration of the original rock. What appears to have been feldspar crystals and lapilli fragments are now replaced by tourmaline, and in some cases tourmaline bands occur in multiple deformed sheath fold structures. It is unclear whether the tourmaline itself has undergone this deformation, or if it has replaced minerals in a pre-existing structure. Thin quartz veins that cut the schist also show varying degrees of deformation, both brittle and ductile. Gold and copper grades are highly variable in the schist, normally increasing toward the contacts between the schist and the breccia. Pyrite/chalcopyrite is up to 25% of the rock mass, with abundant chalcopyrite and molybdenite.

The quartz-tourmaline breccia portion of the Blue Whale exhibits the highest gold and copper grades of the Siembra Minera Project. Tourmaline has completely replaced blocks of the breccia, while quartz has flooded the matrix. This rock does not show the strong ductile deformation of the sericite-pyrite-quartz schist. Chalcopyrite is the dominant sulphide, with lesser pyrite, bornite, covellite, and molybdenite. Other alteration minerals present are sericite, rutile, calcite, albite, siderite, and minor anhydrite (the latter occurring in undeformed, crosscutting veinlets).

 

---

www.rpacan.com

DISSEMINATED GOLD+PYRITE± COPPER

The bulk of mineralization occurs in disseminated, coalescing, lensoid bodies, and high in gold and in most cases low in copper. These bodies lie almost exclusively in the lapilli-rich, rapidly alternating sequence of tuffaceous units and are clearly aligned along foliation. Together, these lenses form a generally well defined mineralized band, which mimics the dip of the foliation/bedding and remains open at depth. It remains at a similar thickness from the northern concession boundary for a distance of 1.4 km south, after which, it tapers rapidly. Alteration minerals characteristic of these lenses are epidote, chlorite, secondary biotite, and sericite.

The gold in the stratiform lenses is highly disseminated but only roughly associated with high occurrences of pyrite. Fine-scale sub-sampling of three metre assay intervals indicates good correlation between gold and small (<1 cm) calcite/quartz veins. Correlation also exists with zones of high occurrence of epidote, and in lapilli-sized lithic fragments that have been partially to completely replaced by epidote and sulphides. Sub-sampling evidence also suggests that gold is more evenly distributed through the rock near the center of the large mineralized lenses than it is near the margins. In section, east-west contours of gold grades at 0.75 g/t or 1.0 g/t show a geometry that essentially mimics contours drawn at 0.40 g/t.

DISSEMINATED HIGH COPPER/LOW GOLD

Stratiform lenses of high copper with or without high gold underlie the gold+pyrite lenses described above. These lenses outcrop in the northern part of the deposit, and plunge to the south in a manner similar to the Blue Whale and high gold/low copper lenses but with variable dips. Deep drilling has intersected these lenses as far south as 681,900 N. Within the stratigraphic column, these lenses generally occupy the TV and TVC-M units. Rock in the mineralized zones is characterized by a high degree of lapilli and crystal replacement by chalcopyrite, and in some cases, by bornite and covellite. High chalcopyrite in the rock matrix is often accompanied by high chlorite, secondary biotite, and in some cases molybdenite.

GOLD-BEARING SHEAR ZONES

Shear-hosted gold occurrences exist in the southern part of the concession, running parallel to the foliation as with mineralization further north. Stratigraphically, they occur above the large disseminated lenses previously described. The gold and copper grades are erratic and discontinuous.

 

---

www.rpacan.com

MINERALIZATION AT CRISTINAS

The main two styles of mineralization present at Cristinas are:

1.	Stratiform mineralization at Conductora, Morrocoy, and Cordova.
2.	Hydrothermal breccia-hosted mineralization at Mesones-Sofia.

STRATIFORM MINERALIZATION

The Conductora (including Cuatro Muertos and Potaso), Morrocoy, and Cordova areas contain over 95% of the gold resource at Cristinas. Mineralization in these zones (here called Conductora-style mineralization) is stratiform in nature and is concentrated in volcaniclastic units within the mafic-to intermediate-composition volcaniclastic host sequence. The distribution of mineralization is controlled by the permeability of the host rocks; gold grade and alteration intensity typically decrease abruptly at the contact between permeable volcaniclastic units and impermeable lava layers, for example. Pre-mineralization, altered dioritic intrusive stocks are largely devoid of significant gold mineralization due to their low permeability.

Mineralization occurs in a greater than three-kilometre long, north-trending zone that dips moderately (30° to 40°) to the west, sub-parallel to the volcanic stratigraphy and to the pervasive (S1) cleavage. Gold mineralization is associated with a sulphide assemblage that consists essentially of pyrite and chalcopyrite.

Alteration mineral assemblages in Conductora are secondary biotite, minor potassium feldspar, calcite, chlorite and minor epidote and sericite. Calcite is ubiquitous, occurring mainly as disseminations, in addition; in carbonate-sulphide veinlets, carbonate-only veinlets, and quartz-carbonate veinlets. Silicification is minimal in Conductora-type mineralization. Minor tourmaline disseminations occur in some parts of Conductora, but in much lower concentrations than in the Mesones-Sofia area. The most consistent gold mineralization occurs in zones in which secondary biotite is most intensely developed. Many sulphide clots within these biotite-dominated alteration zones are rimmed by a green chlorite alteration that has overprinted the secondary biotite.

Pyrite and chalcopyrite constitute the only sulphide species of significance in primary ore. The average pyrite/chalcopyrite ratio is greater than five. Sulphides occur principally as disseminations, but also in narrow veinlets 1 mm to 2 mm wide. These veinlets are variable

 

---

www.rpacan.com

in composition ranging from sulphide-only to sulphide-calcite and sulphide-calcite-quartz.

These veins have selvages of secondary biotite, chlorite, or chlorite-epidote.

Quartz-sulphide veins are rare, but where they do occur, they are in zones of intense secondary biotite development against which they have indistinct margins and are associated with multi-ounce gold values. Higher than average gold grades (>2 g/t) are associated with areas in which pyrite occurs as coarse clots up to 2 cm in diameter in zones of intense secondary biotite alteration. Generally, however, the sulphides are fine-grained, and much more so than in Mesones-Sofia.

Molybdenite is locally quite abundant, occurring in quartz-calcite-sulphide veinlets, and disseminated with pyrite and chalcopyrite. The Potaso area contains disseminated molybdenite that appears to have no spatial relationship with pyrite and chalcopyrite on a hand-specimen scale.

QUARTZ-TOURMALINE-SULPHIDE-CALCITE VEIN BRECCIAS

Mineralization in Mesones-Sofia is concentrated in the quartz-tourmaline-sulphide-calcite vein breccias and extends laterally into the adjacent country rocks. The breccias are sufficiently closely spaced that the country rock between them also constitutes ore in the central part of Mesones-Sofia. Grades in the country rock on the periphery of the system decrease as the distance between the breccias increase.

Breccias consist of quartz, tourmaline, calcite, and sulphides, and the country rock alteration assemblage consists of fine-grained quartz, muscovite (sericite), calcite, tourmaline, and disseminated clots of sulphides. Silicification is variably developed, with pervasive silicification largely confined to the breccias where it encapsulates the sulphides. Muscovite gives way to secondary biotite in the deepest intercepts in Mesones-Sofia. The occurrence of relict laths of biotite within intensely sericitized zones, as well as relict biotite in the central parts of larger muscovite laths, provides evidence that muscovite replaced pre-existing secondary biotite in the upper parts of the Mesones-Sofia hydrothermal breccia system. Patchy potassium-feldspar alteration is evident in the central part of Mesones-Sofia.

Sulphides commonly occur in aggregates up to 5 cm in diameter at Mesones-Sofia. Sulphides also occur as semi-massive replacements in the matrix of the quartz-tourmaline breccias and as disseminations both in the breccias (in the matrix and in breccia clasts) and in the enclosing

 

---

www.rpacan.com

country rocks. Sulphide content in primary, hard-rock ore is 5% to10% with a pyrite/chalcopyrite ratio of less than 5. Pyrite and chalcopyrite are the only common sulphides in Mesones-Sofia; molybdenite is scarce, but where it does occur, it is associated with pyrite and chalcopyrite. There is evidence that chalcopyrite gives way to pyrite upwards in the breccia bodies. For example, breccia bodies at Morrocoy, located structurally 200 m above Mesones- Sofia, have similar overall sulphide contents but contain only a minor proportion of chalcopyrite. There is no appreciable difference in the nature and distribution of sulphides, sulphide species, or grade, between the muscovite- and biotite-dominated alteration assemblages. This implies that the majority of the mineralization was in place by the time that secondary biotite was overprinted by muscovite.

OTHER MINERALIZATION

Discrete auriferous quartz veins are located adjacent to the Cristinas deposit. Such veins include the Los Rojas and Albino veins that lie approximately one kilometre to the east of the Conductora area, and the Hofman vein, which lies about one kilometre to the west of the Cordova area. These veins consist of quartz with gold mineralization associated with pyrite (there is no appreciable chalcopyrite). The veins have chlorite selvages about 50 cm wide. Although gold mineralization in these veins does not constitute part of the Cristinas resource, they are considered to be genetically related, and peripheral, to the Cristinas deposit. This type of mineralization is not discussed further in this report.

 

---

www.rpacan.com

8 DEPOSIT TYPES

BRISAS

The Blue Whale body has been interpreted to be structural feature, a dilational zone of weakness that has acted, at some point after deposition of the tuffaceous rocks, as a conduit for mineralizing fluids. Based on structures seen within the Blue Whale, this occurred before or during regional metamorphism. The initial pulse of mineralization probably occurred when the system was relatively young. Brecciation, on a limited scale, took place along a pre-existing fracture with fluids rich in B, Cu, Au, and lesser Mo. Alteration in and directly around this feature was intense, causing complete replacement of breccia fragments by tourmaline, massive quartz, and copper.

A possible deposit analogy is of a copper porphyry forming over a magmatic source (yet to be discovered) that was very rich in boron. A peraluminous granite might fit the boron requirements and a sufficient volume of basaltic/andesitic rock could provide the copper. Thin lenses of high Cu and Mo extending away along bedding/foliation planes could be the result of periodic high confining pressures within the Blue Whale that forced mineralizing fluids outward along these planes. The fluids replaced crystals and lithic fragments, evidence of which can be viewed in drill core.

The bulk of gold mineralization at the Brisas deposit appears to have been emplaced after formation of the Blue Whale mineralization. It occurs over a wide area and the highest gold grades do not occur in proximity to the Blue Whale. Although on a small scale gold appears to link with zones of higher schistosity and development of alteration minerals, on a larger scale it was deposited in favourable lithologic hosts, comprising mostly thin and variable tuffaceous rocks. Improved permeability related to bedding discontinuities and relatively porous lithic fragments may have been the overriding factor in mineralization deposition. The fluid pressures must have been high to disseminate them through an unfractured volcanic pile rather than along obvious shear planes or fractures. Mineralogically, this phase of deposition bears some similarity to the high temperature B, Cu, Au, and Mo fluid phase proposed for the Blue Whale, specifically in regards to the formation of disseminated lenses. Geometrically, this package of lenses plunges to the south, where it can still be detected by deep drilling. This pattern is similar to what is observed in the Blue Whale.

 

---

www.rpacan.com

CRISTINAS

In terms of classification, Cristinas has been assigned to shear zone-hosted systems by some geologists, and to a porphyry association by others; however, several key elements of the Cristinas deposit must be satisfied in any attempt to classify the deposit. These include:

• Hydrothermal quartz-tourmaline breccias are present at the core of the mineralized system. Although the quartz-tourmaline breccias contain less than 5% of the Mineral Resource at Cristinas, the concentric arrangement of alteration facies record a decrease in hydrothermal fluid temperatures away from the breccias, showing that these constituted the core of the hydrothermal system. The quartz-tourmaline breccias are features that cross-cut stratigraphy and are clearly not shear zone related. Similar breccias are common in porphyry environments, and some show alteration and metal zoning similar to that observed at Mesones-Sofia, such as secondary biotite at depth to quartz-sericite at shallower levels, as well as a decrease in chalcopyrite upwards within the breccias.
• Alteration zoning at Cristinas is similar to that associated with porphyry systems with proximal secondary biotite giving way to distal chlorite-epidote-calcite (propylitic) facies. Quartz-sericite alteration is superimposed on other alteration facies and is likely to have resulted from the draw-down of meteoric water as the prograde hydrothermal system collapsed.
• The metal association of gold with copper and minor molybdenum is similar to porphyry systems and is not common in shear zone-related gold systems.

Despite these factors that are typical of porphyries, Cristinas clearly is not a classic porphyry system, since mineralization is not contained within, or closely associated with, any porphyritic intrusive stock. Furthermore, the abundant quartz veining associated with most porphyries is largely absent from the Cristinas deposit.

 

---

www.rpacan.com

9 EXPLORATION

The history of exploration work completed on the Project is described in Section 6 History.

EXPLORATION POTENTIAL

The Siembra Minera mineralization is open down dip in all zones and along strike to the northwest in Morrocoy and Cordova because of insufficient drilling. Current resource pit shells are limited by drilling instead of economics. Current plans for exploration are based on brownfield expansion of the existing deposit. RPA is of the opinion that there is excellent potential to increase the resources and to convert a significant portion of the Inferred Mineral Resources to Indicated with more drilling. RPA recommends drilling approximately 150 to 200 drill holes totalling approximately 75 km to 100 km. This drilling would have a number of objectives including:

• Conversion of Inferred Mineral Resources to Indicated with priority set on Inferred Mineral Resources situated in the 5 and 10 year pit shells.
• Drilling to determine the extent of mineralization at depth in the Main Zone as this will determine the limits of the largest possible pit and help with the location of features such as dumps and roads.
• Better definition of the copper mineralization in the Main Zone footwall.
• Improving preliminary artisanal mining sterilization assumptions.
• Condemnation drilling of proposed waste rock storage sites.
• Closer spaced drilling in the El Potaso area between Brisas and Cristinas.
• Drilling on the northwest extensions of the mineralization in the Morrocoy and Cordova areas.
• Drilling on the Cristinas Main Zone for density measurements.

Figure 9-1 illustrates some of the exploration targets. Most of the drill holes target down dip extensions of the Main Zone and the boundary area between the Brisas and Cristinas concessions known as Potaso. The average length of these holes has been estimated to be approximately 500 m. An approximate cost for this drilling ranges from approximately $15 million to $20 million.

 

---

---

www.rpacan.com

10 DRILLING

GENERAL

GRI began exploration activity on the Brisas concessions in late 1992, and continued with various drilling programs through 2006. A total of 975 drill holes with a total drilled length of 207,442 m have been completed by GRI on the Brisas concessions as of September 2006 (Table 10-1). Of these holes, 802 representing 189,985 m of drilling were completed specifically for exploration. The remaining holes were drilled for hydrologic, geotechnical, and metallurgical testing, and in some cases, were assayed and used in modelling.

Drill holes within and around the planned pit area are mostly spaced at approximately 50 m or 100 m apart. Drill hole spacing in the Blue Whale area is approximately 25 m. The majority of the exploration drilling was performed using standard diamond core-barrel recovery techniques although a small amount of auger drilling was carried out at the beginning of the exploration campaign. Auger holes (“A” holes) are generally very shallow, located throughout the Project area and in particular between later-drilled core holes; many auger holes are outside the pit area.

Placer drilled 1,182 holes on the Cristinas concessions between 1994 and 1997 (Table 10-2). Once early exploration drilling determined the approximate location and strike direction of mineralization, further drilling was undertaken on section lines orientated perpendicular to that trend. The drilling completed in the southern two-thirds of the Cristinas concessions has shown that the mineralization occurs in a large tabular body that strikes approximately north-south, and dips moderately to the west. The drilling completed in the northern third of the Cristinas concessions has shown that the strike of the mineralization has changed in this area. In this northern portion of the Cristinas concessions, the mineralization can occur as pipe-shaped forms, and as thinner tabular forms with sub-vertical dips.

 

---

www.rpacan.com

TABLE 10-1 SUMMARY OF GRI DRILLING-BRISAS CONCESSIONS

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

 

---

www.rpacan.com

Crystallex drilled 90 holes on the Cristinas concessions from 2003 to 2007. The information from these drill holes were used by MDA in 2007 to prepare the update of the Mineral Resource estimate and were discussed in summary form in MDA's accompanying Technical Report (MDA, 2007). While the significant intersections from these drill holes were reported in news releases, insufficient details regarding the exact location and inclination of the Crystallex drill holes or the individual assay results were presented in these news releases to be useful. As the results from this drilling campaign were not available to RPA and the information could not be reconstituted from the news releases, none of the drill holes completed by Crystallex were used in preparing the current Mineral Resource estimate.

The drill hole locations are shown in Figure 10-1.

 

---

Mesones W

Cordova ^Morrocoy Mesones E

Cristinas

10-4

---

www.rpacan.com

BRISAS CONCESSIONS

The following is taken from PAH (2005 and 2008).

COLLAR AND DOWNHOLE SURVEYING

For all auger and core holes at the Brisas Project, the field location of the drill hole collars before drilling and collar surveying after drilling have been performed by Surco, a local survey firm based in El Callao, Venezuela. This company was also responsible for establishing the concession boundaries and setting up permanent survey reference points within the concession. The base for all surveys was GPS defined and checked by the survey company with a traverse from a nearby GPS station (Las Cristinas) with satisfactory accuracy.

The setting up of the bearing and inclination of the drill rig was made with compass and inclinometer. All core holes were surveyed with a Sperry Sun photographic instrument mounted inside a rod that can be inserted into the drill hole using the drill equipment, recording azimuth and dip at varying depths by technicians employed by GRI. The first photo was normally taken at a depth of approximately 20 m (without casing), a second photo at 6 m below the cased intervals (below the saprolitic zone), and subsequent photos every 100 m to 150 m thereafter. The reading on the developed film was checked by a geologist and the information entered into a field book.

CORE LOGGING

The logging format for the Brisas Project had several changes through the different drilling stages as adjustments to Rock Quality Designation (RQD) measurements and standardization of lithologic and alteration codes were made. The code standardization was implemented after drill hole D95, and many of the previous holes were re-logged to avoid differences in log codings. Two different log forms, geotechnical and geological, were completed when logging.

The geotechnical log completed for each hole included depth, bit diameter, core recovery, rock hardness, sampling intervals, and RQD. Core recoveries were generally good averaging approximately 96%. An average recovery of 87% was obtained in saprolite, and 98% in hard rock. The core recovery for the Blue Whale was 91%. RQD, as measured by GRI, is the ratio between the cumulative length of naturally un-fragmented/un-fractured core longer than 0.1 m and the total core length within a 3.0 m standard measurement interval. RQD readings were obtained before sampling and/or destruction of the core and recorded in the logs. Drilling was

 

---

www.rpacan.com

normally performed with HQ diameter (2.5 in. or 6.35 cm) to the saprolite-hard rock contact where bits were changed to NQ diameter (1-7/8 in. or 4.76 cm). Due to the characteristics of saprolite and other intensely weathered rock, RQD readings were not made above the hard rock-saprolite contact.

Detailed geological logs were recorded on a form with the following information:

• hole number, summary of location (coordinates, elevation), drilling company, geologist, and date;
• log description, including rock type, degree of oxidation and weathering, intensity and type of alteration, sulphide mineralization, veining or other structure(s) (jointing, fracturing and brecciation), and rock color. Alteration and mineralization minerals were quantitatively estimated and recorded, as well as all other parameters, for computer input.

A summary log was then completed from the detailed information, along with a graphical interpretation of the log, as well as gold and copper assay results. Logging procedures followed by GRI were well established and have been followed by all geologists, with minor changes, through the different exploration stages. Quality was assured through the use of an internal manual: “Procedures for geological logging at Brisas del Cuyuni”, which provides guidance in the use of geological terms, defines different lithological units, structure and visual evaluation of alteration and mineralization contents.

TWIN DRILLING VERIFICATION

Twin hole tests were run occasionally throughout the Brisas Project drilling program. A total of seven twin holes were drilled at different times and locations within the property. Both the initial and the twin were core holes. A more detailed discussion of twin hole data results is presented in Section 12 of this Technical Report.

CONDEMNATION DRILLING

Condemnation drilling has been performed extensively on the Brisas concessions. Both condemnation and geotechnical drilling has been performed on the proposed waste dump areas and plant site. Geotechnical drilling was conducted on the proposed tailings dam area for which some assay information was obtained. None of the drilling of these areas has yielded geological or geochemical information suggestive of the presence of significant mineralization, and therefore no additional condemnation drilling was recommended for Brisas by PAH (2008).

 

---

www.rpacan.com

RPA recommends investigating if additional condemnation drilling will be required for the Siembra Minera Project.

SAMPLING

In auger drilling, each three-metre auger flight was lifted onto a table and the soft saprolite was peeled off, dried, and prepared for assaying. In core drilling the soft saprolite was cut longitudinally by machete and the hard rock core cut by a standard Clipper 12-inch diamond saw. Half of the core was placed back in the core box for storage while the other half was placed in metal trays for drying in a fuel oil boiler for sample preparation. Most core drilling was done with HQ core (63.5 mm diameter) but deeper holes were sometimes reduced to NQ (47.6 mm diameter) to accommodate the depth capacity of the drill rig.

GRI maintained a full record of split core for the entire drill program. The sampling interval was generally three metres, with the exception of samples adjacent to the saprolite-hard rock contact, where in some cases adjustments were made to differentiate sample types, or in a few holes located in exploration areas outside the main mineralized zone (e.g., D722-D727), where a one-metre interval was used. The sample size was nominally eight kilograms in weight for the three-metre sample.

The gold and copper mineralization at the Siembra Minera Project is broadly disseminated and amenable to bulk mining. The deposit is proposed to be mined on six-metre benches in ore zones and 12 m benches in waste zones. In RPA’s opinion, the three-metre sample length is adequate and generally provides sufficient resolution in defining the ore and waste zone boundaries for the mineralization except perhaps for the Blue Whale zone, which tends to be narrower and of higher grade than the rest of the deposit and hence, a shorter interval may have allowed for better boundary definition. On the other hand, the longer interval will tend to incorporate some dilution to the model.

RPA is of the opinion that the drilling, sampling and logging procedures carried out on the Brisas concessions meet industry standards, and are suitable for use in the preparation of Mineral Resource estimates.

 

---

www.rpacan.com

CRISTINAS CONCESSIONS

The following is taken from MDA (2007).

COLLAR AND DOWNHOLE SURVEYING

According to historic Placer documentation of the drilling procedures, drill hole locations were established using a prismatic or Brunton compass, and adjusted into position with a Brunton compass. After completion, each hole was fitted with a collar pipe, and a cement collar block was inscribed with the drill hole number. Final drill hole collar locations were then surveyed in UTM coordinates by Surco, translated into local grid coordinates, and entered into a GEOLOG database (a proprietary drill hole database format). Examination of the drill hole deviation measurements shows that 907 of the 1,174 holes (77%) have at least one downhole survey. Downhole survey readings, generally taken approximately every 50 m, were completed using a Sperry Sun single-shot survey camera or a Pajari compass. The GEOLOG database contains detailed geological descriptions, geological codes, check assay data, specific gravity data, core recovery, RQD data, and some trace element geochemical data.

CORE LOGGING

Drill core was logged for rock type, alteration, mineralization, structure, and magnetic susceptibility. In addition, RQD, core recovery, rock strength, and joint roughness and coating were logged. If core recovery in the saprolite averaged less than 80%, the hole was re-drilled at the contractor’s expense; global average core recovery in saprolite was between 85% and 90%. Hard-rock core recovery was above 95%. Oriented core was drilled in selected areas using a downhole “crayon test” for determining the true orientation of foliation, bedding and lineation, as well as the orientation of veins and veinlets.

SAMPLING

Drilling in an intensely weathered tropical environment presented challenges and, consequently, several different drilling techniques were attempted by Placer before choosing triple tube diamond drilling. Other methods tested include Vibracore, auger, and reverse circulation rotary drilling, none of which produced acceptable results. Up to seven hydraulic diamond drill rigs were used simultaneously to complete the drilling. The best recovery was achieved with PQ tools (85 mm diameter) in saprolite, and with HQ tools (61 mm diameter) in bedrock. HQ was also used to drill some of the saprolite, as not all rigs were equipped to

 

---

www.rpacan.com

handle PQ (85 mm diameter) core. NQ (47.6 mm diameter) was used systematically in bedrock during the infill drilling phase within the Stage I pit area and occasionally in difficult drilling situations. The saprolite interval was drilled uncased until casing could be set in bedrock. Sample lengths ranged from 0.1 m to 8.0 m, with most being approximately one metre.

RPA is of the opinion that the drilling, sampling, and logging procedures carried out on the Cristinas concessions meet industry standards and are suitable for use in the preparation of Mineral Resource estimates.

 

---

www.rpacan.com

11 SAMPLE PREPARATION, ANALYSES AND SECURITY

BRISAS CONCESSIONS

The following is taken from PAH (2005).

SAMPLE PREPARATION

Sample preparation, including drying, crushing and pulverizing, was performed on site at GRI’s own sample preparation facility using the sample preparation routine summarized in Figure 11-1. The sample pulps were shipped to assay laboratories in Puerto Ordaz, Monitor Geochemical Laboratory de Venezuela, C.A. (Monitor) and Triad Laboratory (Triad), during the earlier campaigns before 1999. The Triad laboratory located at Minera Hecla’s La Camorra mine site was used for the later round of drilling (2003-2004). After drying, all samples were crushed to 90% –8 mesh (2.36 mm). Half of the crushed sample was bagged and sorted for reference; and a split of approximately 500 g from each sample was pulverized to 90% –150 mesh (0.106 mm). Crushing was carried out with 6x4-inch Morse and 4x8-inch Marcy jaw crushers and a roller crusher. Pulverizing was accomplished with Bico puck and ring pulverizers, although Bico disk pulverizers were also available. Pulverizer cleaning with barren sand was performed after every ten samples. Quality assurance/quality control (QA/QC) procedures included sending pulps to Acme Labs in Vancouver for checking one of every 20 samples and inserting standards prepared with the Brisas mineralized material by Hazen Labs at a rate of 1 in every 30 samples.

Assay laboratories used during the early stages of the Brisas Project drilling were Barringer Research Labs (Barringer) and Bondar Clegg Labs (Bondar Clegg). Monitor was used as the primary assay laboratory and Triad was used as the check assay laboratory from 1994 to 1999 when checks established confidence for these two local laboratories based then in Puerto Ordaz, Venezuela. For the 2003-2004 drilling program, Triad located at Minera Hecla’s La Camorra mine site was used. During that time period, Triad worked with Acme Labs in Vancouver, Canada, for check assaying purposes and also participated in international round robin assaying programs.

 

---

www.rpacan.com

For the 2003-2004 drilling program, samples were prepared on site and pulps sent to Triad. The laboratory routinely ran samples for Hecla’s La Camorra mine where it was located, as well as for other companies operating in Venezuela. Assaying control procedures included log record and tag identification of samples, a control list, blank and rejects run on approximately 10% of samples, assay check runs on approximately one of every 15 samples. Both gold and copper assaying were performed using standard fire assay (FA) and atomic absorption (AA) techniques for the Brisas Project. RPA toured the Triad laboratory during a visit to La Camorra in May 2005 and found it to be reasonably well operated. Triad had begun the accreditation process and participated in bi-annual round robin assaying of Geostats reference standards. In 2005, Triad was registered in Arizona as a fire assay laboratory and had operated in Venezuela for twelve years, including the past five years at the La Camorra mine site.

 

---

www.	rpacan.com

11-3

---

www.rpacan.com

ANALYTICAL METHODS

Analytical methods used for the early stage of the drilling programs were metallic screen analysis for gold and geochemical analysis for copper. During 1994 to 1999, all pulp samples were analyzed for gold by FA with an AA finish. Samples with gold values over 1.5 g/t Au were re-assayed with 1.0 assay ton FA with a gravimetric finish. Copper assay was performed using standard AA with long iodide titration verification when values were obtained above 0.3% Cu.

QUALITY ASSURANCE AND QUALITY CONTROL

The Siembra Minera Project generated a large amount of assay information consisting of original assays, checks, and standards that were routinely received. These data were kept both in original hardcopy and in digital format. Assays were checked for correct sample number, intervals, actual values from the laboratories, and finally for conflicts within the primary laboratory, and between the primary laboratory and the check laboratory. If samples had conflicts (i.e., [A-B]/[A+B] > 33% variance), they were reviewed and if necessary the laboratories were requested to re-assay. In some cases, there were up to five check assays for a given sample interval for several high-grade gold assays. For standards, the tolerance was a variance of 12% for both copper and gold. For drill holes with serious standards conflicts, the entire drill hole could be requested to be re-assayed. Once the conflicts were resolved, all assay data were kept in an “Accepted Assays” spreadsheet under the control of the project manager. The analysis of assays through the use of the spreadsheet as a control provided a reliable method of determining conflicts between primary and check laboratories. This method was designed by GRI in 1995 with subsequent audits and modifications by independent parties (Mark Springett 1995, Behre Dolbear 1997). The actual assay value included in the drill hole database and utilized in modelling is the average of all accepted assays for a given sample interval.

DENSITY MEASUREMENTS

From 1994 to 1997, there was an ongoing program totalling hundreds of field measurements of bulk densities and moisture contents. The following methods were used for bulk density measurements:

• Method 1: known volumes were excavated from road or trench cuts and weighed;
• Method 2: boxes of drill core were weighed and adjusted for the box weight and estimated volume of core loss; and

 

---

www.rpacan.com

• Method 3: individual core pieces were weighed and volumes estimated by volume displacement.

Moisture content was measured by weighing new core, drying it overnight, and re-weighing it. In-place, dry bulk densities and moisture content for different rock/alteration types were compiled by GRI for resource/reserve studies based on all valid information using a weighted average method (Table 11-1). The densities were used in the January 2005 feasibility study and subsequent studies and are grouped by rock type and degree of weathering. The main groups are oxide saprolite, sulphide saprolite, and un-weathered rock. An additional category was created for schist, because it consistently had a lower density for un-weathered rock than other rock types that are generally considered as “tuff.”

TABLE 11-1 MATERIAL DENSITIES AND MOISTURE

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

CRISTINAS CONCESSIONS

The following is excerpted from MDA (2007).

SAMPLE PREPARATION

Although sample preparation and analytical procedures are well described in Placer’s reports, it is not clear what special security procedures were in place at that time. The Triad laboratory of Tumeremo, Venezuela, and Bondar Clegg, Vancouver, Canada assayed all samples taken at Las Cristinas in 1992. Beginning in January 1993, Placer Research Centre in Vancouver, Canada, assayed all core samples, while Monitor analyzed trench samples.

All samples were prepared on-site. In 1993, staff from Placer Research Centre reviewed and amended laboratory procedures to conform to Placer standards. Figure 11-2 shows Placer’s sample preparation procedures.

 

---

www.rpacan.com

ANALYTICAL METHODS

All samples were fire assayed for gold and “geochemically” analyzed for silver, molybdenum, copper, and cyanide-soluble copper. Table 11-2 shows the assay techniques used on Las Cristinas samples by the Placer Research Centre. Note that the term “geochem” was not explained.

 

 

Notes:

1.	Au > 3 g/t were re-analyzed with a gravimetric finish.
2.	Ag > 10 g/t were re-analyzed using same analytical procedures.
3.	Cu > 4,000 ppm were re-analyzed using same analytical procedures.
4.	CNSCu is cyanide soluble copper.
5.	Mo > 1,000 ppm were re-analyzed using same analytical procedures.

In addition to the above elements, core samples collected early in the program were analyzed for mercury, antimony, arsenic, zinc, and lead. Multi-element analysis was also performed on 3,700 surface samples. Additional multi-element analyses were completed on five metre downhole composites from ten holes drilled on cross section 9,600N in the Conductora deposit.

 

---

www.rpacan.com

GR Engineering (Barbados), Inc.

Siembra Minera Project
Bolivar State, Venezuela

Sample Preparation Flow Sheet,
Cristinas Concessions

March 2018

Source:Mine Development Associates, 2007.

11-7

---

www.rpacan.com

QUALITY ASSURANCE AND QUALITY CONTROL

In July 1993, R. Mohan Srivastava conducted an inter-laboratory bias analysis to compare Triad, Bondar Clegg, and Placer Research Centre assay results. The study concluded that the Triad results tended to be biased on the low side, while some of the Bondar Clegg results tended to be biased on the high side. Consequently, it was decided to re-assay all Triad and Bondar Clegg samples for gold only at the Placer Research Centre and to use only Placer’s gold assays on drill core for the 1996 resource study.

Standard, duplicates, and blanks were used for quality control of the on-site sample preparation laboratory. For every suite of 20 samples, there was one each of a duplicate, standard, and blank, which were submitted as blind samples to the assay laboratory.

Thirteen standards were prepared by the Placer Research Centre representing a broad range of gold grades from Las Cristinas surface and core material. These were used to monitor accuracy of the assay laboratory as well as to detect potential contamination in sample preparation. Duplicates were taken from a split of the preceding sample and were used to test the precision of the assays and the homogeneity of nugget effect of the samples. Blanks were obtained from a nearby diorite quarry and were used to detect possible contamination during sample preparation as well as to verify sample order.

Standards, replicate samples on the same sample pulp, and blanks were also used for the quality control program for gold assays at the Placer Research Centre. In each suite of 24 samples, one each was a replicate, standard, and a blank. According to Placer, quarterly statistical evaluations of the QA/QC data indicated that Placer’s laboratory produced accurate and precise gold assay results. Results from a geochemical quality control program also indicated that the Placer Research Centre’s geochemical analyses for copper, silver, and molybdenum were highly accurate and precise.

In addition, 10% of the samples were sent to an outside laboratory for an independent check; the laboratory was the IPL laboratory (IPL) of Vancouver, Canada. Of the 5,866 samples analyzed from 1993 to 1995, the two data sets were quite similar with minor differences between the two laboratories especially for gold grades less than 1.0 g/t Au, according to Placer’s 1996 feasibility report. The average inter-laboratory bias appeared to be approximately 5% to 10%, with Placer’s laboratory results being higher than IPL’s. The Placer 1996 feasibility report noted that this grade range was important because the economic cut-

 

---

www.rpacan.com

off for the project is between 0.6 g/t Au and 0.7 g/t Au. That report stated that “It appears that the PDI laboratory is providing more reliable assays of the less than 1.0 g/t Au gold grades than is the IPL laboratory. IPL appears to be understating the gold grade of the less than 1.0 g/t Au grades by about 5% to 10% …. From this analysis the PDI assay results can be considered appropriate for the resource estimation.” MDA (2007) was unable to definitively analyze and compare the samples and check samples to verify Placer’s above conclusion.

QA/QC information was also gathered on assay samples from the trenching program; these samples were assayed by Monitor. The Placer Research Centre helped Monitor implement in-house standards and also completed a check assay program on samples sent to Monitor. A 1995 evaluation indicated that it appeared Monitor’s assays were on average 5% to 10% higher than the expected means of the standards’ values and that Monitor’s mean gold grades were approximately 7% higher than Placer’s mean gold grades on trench samples assayed by both laboratories. Placer’s 1996 feasibility study concluded that “The systematic bias in the Monitor assay results presented above is not thought to have a significant impact on the 1996 Conductora/Cuarto Muertos resource estimate because the trench data are only a small part of the data base used for resource estimation.” A similar check on Monitor’s results from the 1998 trenching program showed that Placer results were approximately 3% lower than the Monitor results.

Monitor also assayed all the Mesones-Sofia drill core from the 1996 drilling, which represents approximately 55% of all the assays in the Mesones-Sofia area. Placer’s 1998 feasibility study reported that, as with the trench samples, Monitor’s drill core assays appeared to be approximately 5% to 10% higher than check assays by Placer Research Centre. This problem was to be studied further, but MDA (2007) was not aware of any further reported conclusions.

Diamond drilling in the intensely weathered environment, i.e., saprolite, presented potential sample bias (Placer used the term “contamination” and considered it similar to that encountered in wet reverse circulation drilling; to be consistent with Placer’s terminology, the same wording will be used here). Crystallex and MDA noted that this was particularly apparent at Mesones-Sofia, where chunks of siliceous or tourmalinized hard rock were floating in the saprolitic clays. During drilling, water flowing around the core could wash out the clays, relatively increasing the amount of hard, possibly better-grade material.

 

---

www.rpacan.com

Placer’s care for this aspect of sampling is reportedly excellent. While great effort was made to eliminate “contamination”, occasional contaminated intervals were unavoidable, according to Placer. Placer stated that “Suspected contaminated intervals greater than 20 cm were sampled and logged as discrete intervals. If the contaminated interval was less than 20 cm the interval was marked and photographed in place and then removed prior to sampling. All sampled intervals were assayed for gold, copper, and molybdenum in order to assess the potential for additional unrecorded downhole contamination on a case by case basis. A total of 831 samples deemed to potentially be contaminated were eliminated from use by coding. The mean grade of these “contaminated” samples is 3.13 g/t Au with a maximum of 29.73 g/t Au. In addition, 32 trench samples deemed to potentially be contaminated were also eliminated from use in estimation”.

MDA evaluated the “contaminated” samples by selecting all samples lying within the area where “contaminated” samples exist. Descriptive statistics were calculated on all “contaminated” and “not-contaminated’ samples. The results showed that there is a large discrepancy in mean grades between the two sets of data for gold, silver, and copper. MDA capped the outlier samples to evaluate if the differences were caused by these few high-grade samples, but the results remained the same. Placer’s elimination of these “contaminated” samples was justified, and MDA (2007) continued with the practice of not using these samples.

REVIEW OF THE QA/QC RESULTS

QA/QC PROCEDURES

QA/QC data for the Cristinas and Brisas deposits have been collected from various sources, however, full documentation for the portion of the deposit located on the Cristinas concessions generally is not available. The following sections outline the available information on the procedures and policies in place during the data collection at both sites, as well as summarizes and analyzes available results.

CRISTINAS CONCESSIONS

Placer geologists included a duplicate, standard and blank sample within every suite of 24 samples submitted for assay. In addition, 10% of the samples were sent to an outside laboratory for an independent check (IPL). MDA (2002) reports that comprehensive reports reviewing the results of the QA/QC data were prepared quarterly, however, these were not available to RPA.

 

---

www.rpacan.com

RPA received a series of text files containing results of QA/QC data integrated during the drill hole programs from 1992 to 1997. Supporting documentation for the QA/QC data is limited, however, Table 11-3 outlines that from 1993 to 1997, Placer was diligent in including blanks, standards, and duplicates in its sample stream.

TABLE 11-3 SUMMARY OF AVAILABLE QA/QC DATA, CRISTINAS

 

Notes:

1.	Expected values and ranges of standards are not available.
2.	The type of duplicate is not specified in the documentation (pulp, reject, coarse, field, check) but is thought to be a pulp duplicate.

As drill hole information from the Crystallex campaigns (2001 to 2006) is not available to RPA and not included in the Mineral Resource database. Summary information from previous reports on the Cristinas Project (MDA, 2002, 2003, and 2007) which outline procedures and results of QA/QC data used to support the Crystallex drilling campaigns are not discussed in this report.

BRISAS CONCESSIONS

Beginning in 1994, one standard sample was inserted for every 20 samples and one check sample was sent to a secondary laboratory for every 10 samples throughout the drilling campaigns, except for 2003-2004 when one in 20 samples were checked and one in 30 samples was a standard. In addition, blank samples were inserted at random to check residual

 

---

www.rpacan.com

contamination (normally one per drill hole and at the end of each sample run). Standards were included in some of the check assay runs.

In 1995, GRI developed a method of identifying potential integrity issues with assay results from outside laboratories, which was audited and modified by independent parties in 1995 and 1997. The method involved routine monitoring of results at Brisas including checking results for:

• Transcription errors;
• Conflict between the primary and secondary laboratories (tolerance for these sample pairs was 33%);
• Standards reporting assay values outside of accepted tolerance levels (12% for both Cu and Au).

In cases where conflicts were identified, assay re-runs were requested and reviewed. Re-runs were extended to the surrounding samples under the discretion of the reviewer. Up to five check assays have been performed using this monitoring system, with the final accepted value represented as an average value of all of the results, following removal of outliers. Potential issues identified in waste material were not always resolved as they did not affect the integrity of the resource database. RPA did not uncover details or results of any coarse or field duplicate programs.

According to PAH (2008), the reliability of assay results was tested throughout the drilling programs including several specific detailed studies by independent parties, including Mr. Mark Springett (1995 and 1996) and Behre Dolbear (1997), all of which indicated a satisfactory level of precision and accuracy.

A summary of available QA/QC data from the Brisas concessions is presented in Table 11-4.

 

---

 

Notes:

1.	Expected values and ranges of standards are not available.
2.	The type of duplicate is not specified in the documentation (pulp, coarse, field) but it is thought by RPA to be a pulp replicate sample.
3.	Despite drilling campaigns, RPA has no record of QA/QC samples taken in 1992 or 2006.

BLANKS

The regular submission of blank material is used to assess contamination during sample preparation and to identify sample numbering errors.

LAS CRISTINAS

A total of 7,515 blanks were included with samples assayed from 1993 to 1997. RPA analyzed the results using an assumed detection limit of 0.01 g/t Au and defined a sample to have failed if it returned a gold value more than ten times the detection limit (>0.1 g/t Au). RPA determined that a total of 249 samples (3%) failed the defined criteria, and observed that these failures were often clustered together temporally (Figure 11-3). RPA is of the opinion that the number of blank failures at Las Cristinas indicates a low degree of sample contamination or sample mix-ups.

 

---

www.rpacan.com

 

BRISAS CONCESSIONS

A total of 466 blanks were included in the sample stream at Brisas from 1995 to 1999. RPA analyzed the results using an assumed detection limit of 0.01 g/t Au and defined a sample to have failed if it returned a gold value more than ten times the detection limit (>0.1 g/t Au). Individual laboratory performance is listed in Table 11-5 and shown graphically in Figure 11-4.

TABLE 11-5 SUMMARY OF BLANK SAMPLE ANALYSIS, BRISAS

CONCESSIONS

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

 

---

MN - Monitor Laboratory; TL - Triad Laboratory

CERTIFIED REFERENCE MATERIAL

Results of the regular submission of certified reference material (CRM) or reference material (standards) are used to identify problems with specific sample batches, and biases associated with the primary assay laboratory.

LAS CRISTINAS

RPA has compiled an incomplete set of QA/QC samples for the Cristinas deposit, however, the data is missing for several important components. With reference to the standard information, the expected values and ranges for the 17 individual standards reviewed by RPA were not available, limiting the ability to assess bias with the primary laboratory and identify issues with sample batches. It is also unclear whether the standard names (STD-1, etc.) could be referenced to the nomenclature employed in previous reports on the deposit. Table 11-6 summarizes the raw data available to RPA, including the approximate time frame of use of each standard, based on associated drill hole ID.

 

---

 

BRISAS

For much of the exploration drill programs, outlying standard values were considered less important than differences between primary assays and duplicate/repeat/check assays. Any standard differing by over 12% of the original standard value was flagged for further evaluation of the sample batch through re-checking, except in cases where:

1.	The standard was inserted within waste material; or
2.	Surrounding check assays showed agreement with the primary assays.

In total, 11% of standards fell outside acceptable limits, and one percent of these were followed up through re-assaying of the standard and shouldering samples.

RPA reviewed graphs compiled internally by GRI for six gold standards and seven copper standards (of a total of 21 standards included). Results were assessed temporally and by laboratory for bias and trends. Table 11-7 lists the expected value and acceptable range of gold and copper values for the standards inserted within the sample stream at Brisas. Gold standards were prepared by Hazen and Cone Laboratories, but their certification and matrix

 

---

www.rpacan.com

are unknown to RPA. The results from Monitor indicate a higher degree of precision at all grade ranges than Triad.

A large scale mislabelling of standards caused STD. 1, STD. 2 and STD. 3 to have different reference values pre and post drill hole D627. A review of results by GRI also caused the accepted value to change in each of these standards. The post 1997 accepted value and label are maintained in the table.

TABLE 11-7 EXPECTED VALUE AND ACCEPTED RANGE OF STANDARD

MATERIAL, BRISAS CONCESSIONS

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

RPA plotted the results of gold standard STD 1 (pre and post 1997) with time for the Monitor and Triad laboratories (Figure 11-5). In many cases, the standard was assayed twice and the data provided and the plotted results shown represent averages of the two results. Both laboratories show a low bias compared to the accepted value, with Monitor showing good precision and Triad showing only poor precision, which improved in the second campaign of samples.

RPA also plotted the results of copper standard STD 1Y with time for the Monitor and Triad laboratories (Figure 11-6). In some cases (approximately 15%), the standard was assayed twice and the plotted result represents an average of the two results. Both laboratories show a low bias compared to the accepted value, with Monitor showing good precision and Triad showing very poor precision with a potentially significant low bias.

 

---

www.rpacan.com

RPA is of the opinion that results indicate good precision at Monitor, and recommends that future drill programs incorporate three or four matrix matched CRMs that approximate the gold cut-off grade, average grade, and high grades at the Project, and which include a relevant copper component.

The precision and bias observed at both laboratories is consistent with the observations of all other standards and grade ranges reviewed. Without the results of a round robin analysis, it is difficult to assess whether the observed bias is a result of laboratory procedures or whether a revision of the accepted value is warranted.

FIGURE 11-5 CONTROL CHART OF GOLD STD – 1Y

 

---

www.rpacan.com

FIGURE 11-6 CONTROL CHART OF COPPER STD – 1Y

FIELD DUPLICATE SAMPLES, COARSE REJECT DUPLICATES, AND PULP DUPLICATES

LAS CRISTINAS

RPA did not uncover details or results of any coarse or field duplicate programs, and has assumed the available duplicate results are based on a program of pulp duplicates. This assumption is based on a survey of internal Placer reports, and RPA’s understanding of the standard practice at the time the programs were undertaken. Duplicate data was not flagged with a laboratory identifier and therefore RPA’s analysis does not comment on the precision of any specific laboratory. The correlation coefficient of the 7,792 duplicate pairs collected at the Project is 0.95, indicating good correlation for pulp duplicate samples. A total of 30% of sample pairs which had an average grade greater than 0.1 g/t Au plotted outside the expected error margin of ±20%. This is considered by RPA to be a high proportion for a pulp duplicate program but not unusual for gold mineralization. A scatter plot of duplicate sample pairs for gold is shown in Figure 11-7.

 

---

www.rpacan.com

 

BRISAS

RPA did not uncover details or results of any coarse or field duplicate programs, however, GRI conducted a very high number of repeat, duplicate and check assays, up to six per sample. The final value in the Mineral Resource database represents an average grade of the repeated samples, which may represent assays with AAS and/or gravimetric finish and/or assays from different laboratories, with outliers reviewed and removed on a case by case basis by GRI geological department. In order to assess the precision of the primary laboratories employed throughout the drilling campaigns, RPA compiled and reviewed the initial and first pulp duplicate assay (AAS finish only) using basic comparative statistics (Table 11-8), scatter plots (Figure 11-8), and quantile-quantile (QQ) plots. Subsequent duplicate results on any single assay were excluded from the analysis.

 

---

 

 

FIGURE 11-8 SCATTER PLOT OF PULP DUPLICATE SAMPLES AT BRISAS

BC - Bondar Clegg: MN – Monitor; TL - Triad

RPA is of the opinion that the pulp duplicate gold assay results indicate good precision at all three laboratories. There is significantly less scatter at Brisas compared to Cristinas suggesting that the gold mineralization at Cristinas is different with more coarser gold grains.

 

---

www.rpacan.com

EXTERNAL LABORATORY CHECK ASSAYS

LAS CRISTINAS

Internal documentation of procedures during the Cristinas drilling campaigns document that 10% of the samples taken were sent to an outside laboratory (IPL) for an independent check. RPA has not received the results of the check assay program and is unable to comment on any observed bias of the primary laboratory as a result.

BRISAS

As part of GRI’s extensive check assay program, pulp samples from primary laboratories at various periods throughout the drilling campaigns from 1993 to 2005 (Bondar Clegg, Barringer, Monitor, and Triad) have been sent to other outside laboratories to assess bias at the primary laboratory. In addition to the laboratories mentioned above, pulp samples were sent to up to ten other external laboratories; sometimes the same pulp sample was sent to three or even four additional laboratories.

In an effort to simplify the results, RPA has limited analysis to the most prolific primary laboratories, Monitor and Triad, and one external laboratory. Figure 11-9 presents a quantile-quantile plot of the check assay pairs from Monitor and Triad (5,512 pairs). The results indicate negligible bias below 2.0 g/t Au, and a positive bias in favour of Monitor at higher grades. The correlation coefficient between the sample pairs was 93% and this bias was also visible in the scatter plot (not shown). The result may be partially explained by Monitor’s more prolific use of gravimetric finish for higher grade results (2.3% of assays were repeated with a gravimetric finish at Monitor vs. 0.7% at Triad), but it also reinforces the findings of the standards analysis, where Triad results were consistently more scattered and overall lower grade than Monitor.

 

---

www.rpacan.com

 

ML – Monitor; TL - Triad

RPA plotted 435 check assay pairs from Triad versus ActLabs in Ancaster, Ontario, Canada and found a similar correlation (94%), however, a moderately strong high grade bias of Triad compared to ActLabs was seen below 1 g/t Au. Insufficient sample pairs were available to compare ActLabs results with Monitor. Insufficient sample pairs were available to compare Monitor with any other known laboratory (almost all checks were sent to Triad from Monitor, and the checks that do exist have little supporting information to be able to draw meaningful conclusions).

QA/QC CONCLUSIONS AND RECOMMENDATIONS

RPA makes the following conclusions with regard to the QA/QC monitoring programs in place at Brisas and Cristinas from 1994 to 2005:

• QA/QC procedures in place during the time of data collection exceeded standard practices of the 1990s, however, field duplicates and coarse duplicates appear to not have been taken.

 

---

www.rpacan.com

• Reference material in place at both projects does not appear to have undergone a round robin analysis, and the standards were not certified. Expected values of the standards were modified throughout the program and biases observed may be real, or may point toward the necessity of a modification of the expected value.
• Evidenced by a comparison of the Monitor and Triad check assays, and the results of the standard program at Brisas, there is a potential low grade bias at Triad, as compared to Monitor.
• Monitor shows the highest precision of the primary laboratories in place at either project, as evidenced by the high correlation coefficient of pulp duplicates. Both Triad and Bondar Clegg show only moderate repeatability for a pulp duplicate. Considering that the final accepted value in the database has undergone a thorough internal review before being accepted, and that it is often based upon the average of several results, RPA is of the opinion that the pulp duplicate program supports the integrity of the Mineral Resource database.
• Overall, RPA is of the opinion that the QA/QC program was reasonable and the assay results within the database are suitable for use in a Mineral Resource estimate.

RPA makes the following recommendations with regard to the QA/QC monitoring programs:

• Implement a field duplicate sampling program at Brisas at a rate of 1 in 50.
• Implement a coarse duplicate sampling program at Brisas at a rate of 1 in 50.
• Acquire or develop three or four matrix matched CRMs in all future drill programs that approximate the cut-off grade, average grade, and high grades at the Project. Insert at a rate of approximately 1 in 25 and accompany external laboratory check assays at a rate of approximately 1 in 20. Limit the number of laboratories in use at the Project to three and limit the number of CRMs to four.

 

---

www.rpacan.com

12 DATA VERIFICATION

PAH DATA VERIFICATION - BRISAS CONCESSIONS

The following is taken from PAH (2005).

The reliability of assay results was tested throughout the drilling programs including several specific detailed studies by independent parties. Mr. Mark Springett carried out studies on the reliability of sampling and assaying in 1995 and 1996 and concluded that the results from the laboratories (250 samples) showed a satisfactory level of precision and were unbiased relative to each other. Behre Dolbear performed an independent check assay program of 36 samples from six holes in 1997 to check assays produced by Monitor or Triad against results from a third laboratory (Bondar Clegg). Samples were selected with values at different ranges of gold grades. Behre Dolbear’s check assay results showed high correlation coefficients for both gold (0.92) and copper (0.99) and mean values within approximately 5% of each other for both metals.

In 1997, GRI and Behre Dolbear jointly drilled six core holes under Behre Dolbear’s direct supervision and conducted assays independently at different laboratories. Behre Dolbear concluded that procedures utilized to collect assay data met or exceeded industry standards and that the assay results from all laboratories (Bondar Clegg, Monitor, and Triad), were reliable.

PAH conducted several data verifications and validations for the January 2005 feasibility study. PAH visited the Brisas Project facilities, toured the laboratory preparation and core shack areas, and inspected the core and several drill sites during the 2003-2004 drilling campaign in February 2004. PAH visited GRI’s offices in Spokane, Washington to review the original drill hole logs and assay sheets in April 2004.

PAH verified the drill log data and assays against the drill hole database used for the Brisas Project feasibility study. Ten holes located in ten different vertical sections throughout the Brisas Project were checked for collar location, downhole survey, assaying and geological/geotechnical information. Minor discrepancies were found in survey and lithology information between the database and the logs; no errors or discrepancies were found on

 

---

www.rpacan.com

assay information. It was found that several holes in the early stages of the drilling campaigns were not surveyed for downhole deviation (e.g., most AD-series holes and some D-series holes). All AD-series holes were apparently given an average of the deviation observed in the few (approximately 20%) that did have deviation measurements.

The downhole deviation can be up to approximately 40 m on long holes (e.g., AD85 at a depth of 362 m), however, the average depth of the AD holes is 214 m and the average depth of the A holes is 27 m. The number of holes affected is less than 10% of the current database and the area covered has been drilled at closer spacing by later campaigns with deviation measurements. Therefore, the lack of downhole surveying in these holes does not appear to greatly influence the model. Also, auger holes were visually inspected in cross sections and showed generally good agreement with the much more abundant surrounding core hole data.

TWIN DRILLING VERIFICATION

Twin hole tests were run occasionally throughout the drilling program. A total of seven twin holes were drilled at different times and locations within the property. Both the initial and the twin were core holes. Visual inspection of twin drill hole intersects on cross section indicates overall a very good correspondence of mineralized areas in terms of location, length of the zones, and distribution of Au and Cu grades, although the comparison of individual samples shows some variability due to natural deposit local variations (nugget effect).

Table 12-1 shows a summary of the twin hole data. The comparison shows good reproducibility of sampling data, but also suggests consistently lower grades mainly for Au, but also for Cu in the twin or A holes, relative to the original core holes. It should be noted that while this apparent bias may be due, at least partially, to the highly variable distribution of gold within the deposit, it is, in some cases, also the result of having a single very high grade assay skewing the overall average for the hole(s) as seen in Table 12-2, for example for holes D404/D404A and D498/D498A. Without these high assays the results compare much better.

The A holes and a few other holes were drilled in 1999 by GRI under the direct supervision of Behre Dolbear as part of an independent verification of the drilling and assaying programs at the Brisas Project. In order to better understand the apparent bias on the A holes, PAH requested that GRI drill a hole (D754) as a twin hole to one of other Behre Dolbear holes drilled in 1999 (D614). As seen in Table 12-2, the PAH hole returned average grades slightly lower than the Behre Dolbear hole for Au and about the same grade for Cu, indicating that a bias

 

---

www.rpacan.com

more likely does not exist on the sampling and assaying data and as such the twin hole data generally confirm the original assay results.

 

MDA DATA VERIFICATION - CRISTINAS CONCESSIONS

The following is taken from MDA (2007).

 

---

www.rpacan.com

As most of the Las Cristinas database is derived from Placer’s work, it is important to note that based on Placer’s descriptions of their procedures, their data collection and exploration procedures conformed to or exceeded industry standards in effect at the time. If conducted as reported, Placer’s QA/QC program was of high quality. In general, MDA found that, based on reported methodology, Placer’s exploration data were collected in a technically sound manner. According to Placer documentation, quality assurance checks were in place for most of the project, and validation of data was ongoing. Nevertheless, it was clear that additional verification was necessary because one company had completed all development work, there were no independent checks or studies of the work, and most of the original hardcopy data were unavailable for detailed study or auditing.

Under the terms of the September 2002 agreement between Crystallex and CVG, Crystallex obtained an electronic database from CVG, which included Placer’s drill, topographic, geological, and engineering data. At that time, data from 1,174 drill holes and 108 trenches were included in the Las Cristinas database. Although approximately 99% of the drill data were obtained, hard copies of the assay and geological data were not available, leaving a gap in the ability to validate the database.

When MDA visited the Las Cristinas site in October 2002, it found drill pads, drill collars, drill core and samples, core photographs, and other supporting data demonstrating that exploration had been done in a manner not incompatible with what was described in the documentation of Placer’s work. To conduct independent verification, Crystallex drilled 2,198 m in 12 diamond drill holes, for a total of 1,079 core samples, to confirm the presence and tenor of mineralization. These 12 holes twinned previously drilled Placer holes. In addition, 275 QA/QC samples from this drill program were analyzed. The Crystallex drill results and check samples corroborate the general tenor of gold mineralization reported by the previous operator. For additional confirmation, Crystallex re-assayed 262 pre-existing pulps, 200 pre-existing coarse rejects, and 342 quarter-core samples of pre-existing core. Although mean grades are similar for both datasets, there is a large variance in grade between individual pairs of Placer’s core assays and Crystallex’s core check samples. The variance is lower in the pulp and coarse reject checks. As a result of some of these discrepancies, several additional studies were completed to aid in the understanding of grade variability.

Natural grade variability (heterogeneity) is an issue at Las Cristinas. Although it has become better understood through the efforts of Crystallex, it is an issue that should continue to be

 

---

www.rpacan.com

addressed prior to and during production, as it may result in massive misclassifications of ore and waste. The effect of material heterogeneity on the resource estimate will be dominated by local variance and may have instilled a minor low bias to the sample database. The issue is introduced by the distribution of metals originally in primary ore.

For this reason, Pitard (2005) rhetorically questioned: “Can the existing gold grade database, created with diamond drilling and conventional 30-g fire assays, lead to an accurate block model?” To which he responded: “The answer is no. But, with good geology of the various quartz and sulphide events, it can make a world of a difference.” The problem he is referring to is the ability to estimate accurately locally and with precision. MDA believes that this is difficult to do, but the consequence is not so great that it would negatively impact a mine and deposit of this scale in an open-pit scenario; essentially higher grades will be generally where higher grades are estimated to be, and the same with the mid- and low grades. While the gold occurs in the free state, it is generally not coarse grained nor visible but does appear to occur in clots of sulphides. It is not possible to compensate for the issue of a potential low bias instilled in the sample assay results.

RPA AUDIT OF DRILL HOLE DATABASE

SOFTWARE VALIDATION

RPA utilized Surpac’s validation features and Microsoft Excel to check for any errors or potential issues in the drill hole data including:

• Sample length issues;
• Maximum and minimum;
• Negative values;
• Detection limit / Zero values;
• Borehole deviations;
• Gaps;
• Overlaps;
• Drill hole collar versus topography;
• Datum;
• Laboratory certificate versus database values.

RPA did not note any significant errors.

 

---

www.rpacan.com

DRILL HOLE COLLAR VERSUS TOPOGRAPHY

A total of 2,068 surveyed drill hole collars can be compared against the extents of topographic surface. None of the drill holes outside the extents of the topographic surface are part of the resource modelling. Approximately 90% of surveyed drill holes are within -4.0 m and +2.0 m of the topographic surface. A total of 40 (1.9%) of surveyed drill hole collars are more than 5.0 m below the topographic surface, with 16 (0.8%) of surveyed drill hole collars lower than 10.0 m below the topographic surface, and a total of 24 (1.2%) surveyed drill hole collars are more than 5.0 m above the surface, with just three (0.15%) being higher than 10.0 m above the topographic surface: DG792, D854 and DG796 at 21.4 m, 17.8 m and 11.3 m, respectively. None of the drill holes higher than 10.0 m above the topographic surface are located within the mineralization wireframes. Eight of the drill holes lower than 10.0 m below the topographic surface are located within the mineralization wireframes. All eight of the drill holes more than 10.0 m below the surface appear to be located within flooded areas.

A total of 108 trenches can be compared against the extents of topographic surface. Almost all the trench collars (99%) are below the topographic surface and the mean elevation below surface topography is 4.0 m. The collar of four trenches (3.7%) is between 10.0 m and 12.0 m below the topographic surface.

Widespread disturbance of the original surface at the Project has taken place and multiple collar and trench sites are now flooded and their present locations show as flat surfaces at the level of the water table. RPA is of the opinion that the vast majority of the elevation differences between surveyed drill holes and trenches with respect to the up to date topographic surface can be explained by extensive conventional small scale mining activity and disturbance of the original surface and by flooding of former works.

ASSAYS

RPA compared 4% of the sample database to the assay certificates from Triad on the Brisas portion of the Project. No major discrepancies were found.

In RPA’s opinion, the drill hole data is adequate for use in the preparation of Mineral Resource estimates.

 

---

www.rpacan.com

13 MINERAL PROCESSING AND METALLURGICAL TESTING

Due to the advanced level of design achieved for both Brisas and Cristinas previously, a number of metallurgical testing programs were completed on a large number of samples. The summary of metallurgical data is taken primarily from previous Technical Reports that have been filed publicly for the Brisas and Cristinas projects and sections of a feasibility study that was completed by a previous owner for the Cristinas Project (Placer 1996, PAH 2008, and MDA 2007).

For Brisas, GRI completed a feasibility study (PAH, 2005) and basic engineering that were based on a processing flow sheet that included gravity concentration and a flotation concentrator with leaching of the cleaner scavenger tailings to produce doré and copper concentrate. For Cristinas, Crystallex completed a feasibility study that included gravity concentration and whole ore cyanide leaching in a carbon-in-leach (CIL) circuit to produce only doré (MDA, 2007).

These decisions were made because the northern side of the Brisas property contains higher concentrations of copper than the southern side of the Brisas property which has higher gold concentrations. Similarly, the northern side of the Cristinas property contains higher concentrations of copper (Mesones) and the southern side has higher gold concentrations. The data regarding rock type and grade distributions from the new RPA model are summarized in Table 13-1.

 

---

www.rpacan.com

TABLE 13-1 SUMMARY OF RESOURCES AND GRADES BY AREA

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

The relative proportions of material from the historical data is summarized in Table 13-2.

The current plan proposes to process the mined material in both a flotation concentrator and a cyanide leaching facility. The material that contains copper concentrations greater than 0.02% will be processed in the flotation plant and the material that contains lower concentrations of copper will be processed in the cyanide leach plant. For comparison purposes, a summary of the tonnes and grade by rock type from the current mine plan is presented in Table 13-2.

TABLE 13-2 CURRENT SUMMARY OF ROCK TYPES AND GRADES

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

 

---

www.rpacan.com

BRISAS

Four types of material were identified for Brisas. They are:

• Oxide saprolite
• Sulphide saprolite
• North hard rock (High Cu)
• South hard rock (Low Cu)

The two hard rock types (i.e., North and South) are defined based on the copper concentration. North is defined as gold-chalcopyrite-pyrite with a copper concentration greater than 0.05%. South is gold-pyrite with a copper concentration less than 0.05%.

From 1992 to 2005, 20 metallurgical test programs and mineralogical investigations were completed for Brisas. Five pressure oxidation testing programs were completed using copper concentrate. Tailings analysis and characterization programs were also completed.

Grinding test work was completed by MacPherson in 1997. The gross autogenous work index was 21.3 kWh/t and the Bond ball mill work index was 15.4 kWh/t.

Test work conducted by Lakefield Research, Inc. (Lakefield) in 2005 was used as the basis of the feasibility study completed by Aker-Kvaerner and the subsequent detailed design completed by SNC-Lavalin. The testing was completed using all four rock types using the following six groups of samples:

• Oxide saprolite: composite sample obtained from hand-augered samples
• Sulphide saprolite: composite sample obtained from 27 drill hole intervals
• North hard rock: composite sample obtained from 31 drill hole intervals
• South hard rock: composite sample obtained from 29 drill hole intervals
• North hard rock: composite sample obtained from 25 drill hole intervals
• South hard rock: composite sample obtained from 25 drill hole intervals

PAH reported that they considered the samples to be representative of the Brisas deposit. The sample locations are shown in Figure 13-1.

 

---

13-4

4rs - Met. Samples
4rn - Met. Samples
5ox - Met. Samples
5ss - Met. Samples
5rn - Met. Samples
5rs - Met. Samples

March 2018

Source: Gold Reserves Inc., 2017.

 

---

www.rpacan.com

The test work conducted by Lakefield confirmed earlier work which showed that the North samples can produce a marketable copper/gold flotation concentrate but the South samples cannot. As a result, the plan was to blend the ores to produce marketable concentrates. Consequently, the Lakefield tests were conducted using blends of the North and South samples and included gravity separation, batch flotation tests to determine optimum conditions, and eight locked cycle flotation tests (LCTs) that were used as the basis of the process design since they are considered to be a better indication of mill performance than open circuit tests. A summary of the results from the LCTs are provided in Table 13-3.

 

CRISTINAS

Data available for the Cristinas deposit includes a Placer feasibility study completed in 1996 and information from the Cristinas Technical Report that was completed in 2007 (MDA, 2007) which is publicly available. The locations of the metallurgical samples that were used in the two studies are shown in Figure 13-2.

 

---

---

www.rpacan.com

PLACER FEASIBILITY STUDY

In 1996, Placer completed a feasibility study that considered processing by gravity concentration and cyanide leaching for oxide ores and flotation and cyanidation for sulphide ores to produce doré and copper concentrate. The study summarized the metallurgical results from 17 metallurgical test reports.

The rock types identified for Cristinas were:

• Oxide saprolite
• Sulphide saprolite
• Carbonite leached bedrock
• Carbonite stable bedrock

OXIDE SAPROLITE

Placer reported that gold extraction from oxide saprolite samples averaged 94% with a cyanide consumption of 0.30 kg/t and that there did not appear to be a correlation between gold head grade and gold extraction for gold concentrations above the cut-off grade.

Placer also determined that the sulphide saprolite and bedrock ores had high concentrations of cyanide soluble copper and associated high cyanide consumptions. Therefore, the proposed flowsheet for those rock types included gravity concentration, flotation, and cyanide leaching of the scavenger concentrate and the cleaner flotation tailings. The test data indicated that there was no correlation between gold recovery and gold head grade although the recovery did vary by rock type. The test data also established relationships between copper head grade and recovery and copper head grade and flotation concentrate grade for sulphide saprolite and hard rock.

SULPHIDE SAPROLITE

Based on a correlation between copper recovery and the copper flotation feed head grade in g/t Cu, Placer estimated the copper recovery for sulphide saprolite as a function of the copper head grade using the equation:

Cu Recovery = 4.2017 × (��u ��ead, g/t)0.3597

Or:

Cu Recovery = 4.2017 × (��u ��ead, % ∗ 10000)0.3597

 

---

www.rpacan.com

The final copper concentrate grade was also estimated as a function of the copper head grade using the equation:

Concentrate, % Cu = 7.854 + 137.08 × ��u ��ead,% − 248.8 × (��u ��ead)2

Using the test data, Placer also estimated that the gold recovery from sulphide saprolite will be approximately 20% to the gravity concentrate, half of which will be recovered as a table concentrate and the remainder will be processed further. Based on the test data, it was estimated that the gold recovery in the flotation circuit is 46.7%. Placer estimated gold extraction by leaching the cleaner tailings and the second scavenger concentrate using the ratio between the sodium cyanide to copper since the cyanide addition is typically the limiting factor for gold extraction for ores that contain high concentrations of cyanide soluble copper. Placer estimated gold recovery for sulphide saprolite based on pilot plant data. The gold recovery estimates by product for sulphide saprolite are summarized in Table 13-4.

 

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

HARD ROCK

Placer also estimated the copper recovery for carbonate leached and carbonate stable bedrock as a function of copper head grade using the equation:

Cu Recovery = 54.504 + 155.96 × (��u ��ead, %) − 211.19 × (��u ��ead, %)2

The copper concentrate grade is also estimated using the following equation:

Concentrate, % Cu = 30.33 × (��u ��ead, %)0.109

For the bedrock samples, the gold recovery by a Knelson concentrator was found to be an average of 20% of the gold. Placer estimated that half of the gold (i.e., 10%) would be

 

---

www.rpacan.com

recovered as a gravity concentrate from a table concentrate and the other half from the table tailings would be leached with 95% extraction for an additional gold recovery of 9.5%. Gold distribution and recovery in the flotation circuit was estimated by correlating the flotation feed grade to the flotation scavenger tailing grade and assuming that 10% of the mass fed to the flotation circuit would be recovered in the rougher/scavenger flotation circuits. Using these correlations, the scavenger tailings would contain 16.3% of the gold fed to the plant resulting in an overall gold recovery of 83.7%. The gold recovery estimates for hard rock by product are summarized in Table 13-5.

TABLE 13-5 PLACER GOLD RECOVERY ESTIMATE FOR HARD ROCK

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

 

 

 

GR Engineering (Barbados), Inc. – Siembra Minera Project

 

CRISTINAS FEASIBILITY STUDY

According to the Technical Report (MDA, 2007), “In early 2003, Crystallex, SNC-Lavalin, and Goode reviewed available metallurgical test data and performed various trade-off studies.

 

---

www.rpacan.com

These analyses indicated that the production, transportation off-shore, and smelting of a copper-gold flotation concentrate, as proposed by Placer, was a less attractive alternative and that direct leaching of most or all of the mineralized material and on-site production of bullion would give better gold recovery. The trade-off studies also showed that the direct leach process, which is the flowsheet originally selected by Placer, would simplify the process, improve plant operability, and give lower capital and operating costs.”

Crystallex maintained the same rock types as those used by Placer in the earlier work.

Composite samples were composited from individual drill core intervals taken from within the limits of the planned Conductora pit and sent to SGS Lakefield for bench tests and pilot plant tests. A number of the samples were composites containing mixtures of saprolite and hard rock in order to simulate the planned operating conditions for the project.

OXIDE SAPROLITE

The as-received screen analysis for the oxide saprolite sample was 63 µm.

The average gravity recovery for the samples tested at SGS Lakefield was 5.3% at a grind size of 80% passing (P80) 35 µm. Professor André LaPlante conducted his standard gravity-recoverable-gold (GRG) test at McGill University and determined that the oxide saprolite sample contained 39% GRG and concluded that approximately 25% of the gold would be recovered by gravity processing. Subsequently, Knelson used their circuit modelling system and projected a gravity gold recovery between 18% and 20% using the LaPlante data.

Intensive cyanide leaching of gravity concentrates produced from combined oxide saprolite and carbonate stable bedrock samples and mine blend samples produced gold extractions ranging from 95.7% Au to 99.3% Au.

Bottle roll tests (BRTs) were conducted on the tailings from the gravity concentration tests. Initial tests were conducted to investigate the effects of grind size. The results showed little difference between tests conducted at P80 50 µm and P80 75 µm so the 75 µm size was selected. The tests showed that 99% gold extraction (gravity plus leaching) was possible after 36 hours of leaching using the pure oxide saprolite sample. Other tests showed 98% gravity plus leach extraction was achieved across a range of leach times between 24 hours and 36 hours.

 

---

www.rpacan.com

Preg-robbing tests showed values of less than 4%.

Copper leaching from oxide saprolite samples was generally less than 5%. Cyanide consumption was related to the concentration of cyanide soluble copper in the samples, which is highest for the sulphide saprolite samples. The cyanide consumption for samples that were not sulphide saprolite, including oxide saprolite was reported to be between 0.25 kg/t and 0.7 kg/t.

Lime consumption for the oxide saprolite samples was reported to be between 1.0 kg/t and 1.5 kg/t.

SULPHIDE SAPROLITE

The as-received screen analysis for two sulphide saprolite samples was 182 µm and 69 µm, respectively.

The gravity recovery for the sulphide saprolite that was ground to P80 50 µm was 18.4% and the sample that was ground to P80 63 µm was 22.9%.

Samples of pure sulphide saprolite material contained higher concentrations of cyanide soluble copper resulting in lower overall gold extraction and higher cyanide consumption. Combined gravity plus leaching gold extractions ranged between 85% and 94%. Cyanide consumption was found to be correlated with the cyanide soluble copper (CNSCu) concentration. For samples containing less than 370 ppm CNSCu, the total gravity plus leach gold extraction was between 85% and 89% with cyanide additions of 1.7 kg/t to 1.9 kg/t. Lime consumption ranged from approximately 0.4 kg/t to over 1.5 kg/t.

The samples were found to be mildly preg-robbing with 9% and 16% of a 10 ppm spike adsorbed after 24 hours.

HARD ROCK

The gravity recovery of gold ranged from 17.2% to 23.8% for samples ground to P80 54 µm to P80 99 µm with no apparent relationships between grind size and gravity gold recovery or head grade and recovery. LaPlante estimated that the carbonate stable bedrock sample contained 46% GRG. Knelson projected gravity gold recovery between 24% and 27%.

 

---

www.rpacan.com

BRTs resulted in a combined gravity plus leaching total gold extraction between 87.8% and 90.1% with an average of 88.7%, a cyanide consumption of 0.51 kg/t and lime consumptions averaging 0.62 kg/t. Goode reported that there were no apparent relationships between gold head grade and gold extraction or reagent consumptions.

COMBINED SAMPLES

Four samples from Mesones were tested. The combined gravity plus leaching gold extraction ranged from 84% to 88% for grind sizes ranging between P80 71 µm and P80 103 µm. The cyanide additions were between 0.9 kg/t and 1.6 kg/t with an average of 0.77 kg/t and lime consumption average 0.44 kg/t. Goode reported that gold recovery from Mesones would likely be improved with optimization of reagent consumption strategies.

A CIL pilot plant was operated using two different blends of rock types. The first was a blend of 20% oxide saprolite and 80% carbonate stable bedrock, the second was a “mine blend” of oxide saprolite, sulphide saprolite, carbonate stable bedrock, and carbonate leached bedrock. The overall gravity plus leach extraction for the initial blend was 89.6% with a cyanide consumption of 0.7 kg/t and the mine blend sample resulted in 89.3% gold extraction with a cyanide consumption of 0.8 kg/t.

COMMINUTION

Bond Wi and Ai tests were conducted on a limited number of samples. They are reported in Table 13-7.

 

Ball mill Wi values were also estimated using the grinding data obtained when feed was being prepared for the metallurgical tests. The average Wi for tests conducted using blends of all four rock types averaged 13.1 kWh/t. The average Wi for tests conducted using only carbonate stable bedrock samples taken from various depths averaged 16.5 kWh/t. Although tests were

 

---

www.rpacan.com

not conducted to determine the Wi for saprolite samples, the work index was estimated to be between 6.0 kWh/t and 8.5 kWh/t using data from combined samples.

CARBON ELUTION

Two samples of carbon taken from the pilot plant test were eluted using a simulated high-pressure Zadra elution process. The results are presented in Table 13-8.

 

VISCOSITY TESTS

SGS Lakefield measured viscosity using a Haake rheometer. The data showed that 100% oxide saprolite has a critical density of approximately 40% solids at a Yield Stress greater than 8 Pa.

DEWATERING TESTS

Flocculant scoping tests and thickening tests were undertaken by SGS Lakefield and Outokumpu. In general, the best results were achieved with low charge anionic flocculants (e.g., Magnafloc 919). SGS conducted static thickening tests in cylinders without rakes. They determined that the hard rock samples achieved an underflow density of approximately 45% solids by weight with a flocculant dosage of 15 g/t. The unit area required for conventional thickener designs was 0.83 t/m²/h. For sulphide saprolite and a flocculant dosage of 33 g/t, the underflow density was 42% solids with a unit area of 1.04 t/m²/h. Oxide saprolite samples required 23 g/t of flocculant to achieve an underflow density of 42% solids with a unit area of 0.22 t/m²/h.

Outokumpu operated a continuous pilot-scale thickener to conduct 58 tests on nine blends of rock types. For the oxide saprolite samples, the results are similar to the results achieved by SGS Lakefield. The hard rock samples achieved higher underflow densities at lower solids loading rates than the SGS Lakefield tests. The results of the Outokumpu tests show that with

 

---

www.rpacan.com

the correct flocculant, thickener underflow densities of 50% or greater can be achieved at a loading rate of 0.47 t/m²/h with all rock types and blends as long as the oxide saprolite content of the feed is less than 50%.

ENVIRONMENTAL TESTING

Modified US Environmental Protection Agency (EPA) acid base accounting (ABA) tests were conducted on samples from Mesones and Conductora. The tests determined that oxide saprolite samples were classified as non-acid generating, the sulphide saprolite samples may be acid generating, and the acid generating potential (AGP) of other samples was uncertain.

CYANIDE DESTRUCTION TESTS

SGS Lakefield performed natural degradation tests on tailings from the pilot plant test. The sample taken from the test that utilized the oxide saprolite-carbonate stable bedrock blend showed that the weak acid dissociable (WAD) cyanide (CN) concentration dropped from approximately 60 ppm to less than 15 ppm in 55 days. The cyanide concentration in the mine blend samples dropped from less than 110 ppm WAD cyanide to approximately 20 ppm WAD cyanide in 100 days.

Continuous cyanide destruction tests were conducted on the degraded tailing solutions using the sulphur dioxide (SO2) – air process. Acceptable WAD cyanide levels were achieved using SO2 to WAD cyanide ratios of six and three without copper additions.

RESULTS AND CONCLUSIONS

Based on the results of metallurgical testing using Brisas and Cristinas samples, the conceptual processes selected for the combined project include a cyanide leach plant to process oxide saprolite and sulphide saprolite that contains low concentrations of copper to recover gold as doré from gravity concentration and cyanide leaching plus a flotation concentrator to process sulphide saprolite and hard rock that contain higher concentrations of copper. The flotation concentrator will recover copper and gold into a copper flotation concentrate and gold as doré utilizing gravity concentration and cyanide leaching of cleaner scavenger tailings.

RPA compared and combined data from the Brisas and Cristinas projects in order to determine appropriate design criteria for the combined plant. This data was then used to estimate

 

---

www.rpacan.com

equipment sizing and the associated capital costs, estimate recoveries, and estimate operating costs. The Brisas design, equipment sizing, and cost estimates were taken directly from the SNC-Lavalin basic engineering design and cash flows. In general, the design for the oxide leach plant was taken from the Cristinas data.

A summary of the recovery estimates is provided in Table 13-9.

 

 

 

---

www.	rpacan.com

14 MINERAL RESOURCE ESTIMATE

SUMMARY

A Mineral Resource estimate, dated December 31, 2017, was completed by RPA using the Surpac and Leapfrog Geo software packages. Wireframes for geology and mineralization were constructed in Leapfrog Geo based on geology sections, assay results, lithological information, and structural data. Assays were capped to various levels based on exploratory data analysis and then composited to three metre lengths. Wireframes were filled with blocks measuring 10 m by 10 m by 6 m (length, width, height). Block grades were estimated using Inverse Distance (ID) and Nearest Neighbour (NN) interpolation algorithms. Gold and copper grades were estimated into blocks using inverse distance squared and dynamic anisotropy with the Surpac v.6.8 software package. The estimated gold and copper grades were used to calculate NSR values for each mineralized block. Block estimates were validated using industry standard validation techniques. Classification of blocks was based on distance and other criteria.

A summary of the Mineral Resources is provided in Table 14-1. Summaries of the Mineral Resources by material type and mineralized zone are provided in Tables 14-2 and 14-3.

 

Notes:

1.	CIM (2014) definitions were followed for Mineral Resources.
2.	Mineral Resources are estimated at an NSR cut-off value of US$7.20 per tonne for oxide-saprolite material and US$5.00 per tonne for sulphide-saprolite and fresh rock material.
3.	Mineral Resources are constrained by a preliminary pit shell created using the Whittle software package.
4.	Mineral Resources are estimated using a long-term gold price of US$1,300 per ounce, and a copper price of US$3.00 per pound.
5.	Bulk density varies by material type.
6.	Mineral Resources that are not Mineral Reserves do not have demonstrated economic viability.
7.	Numbers may not add due to rounding.

 

---

 

 

---