Bema Gold 6-K Report of Foreign Issuer

Compañia Minera Maricunga

Technical Report on the Refugio Gold Project

Maricunga District, Chile

Prepared for:

Compañia Minera Maricunga

Copiopo Chile

Prepared by:

Kinross Gold Corporation

Technical Services Department

220 Bay St.,

Toronto, ON

M5J 2W4

December, 2003
	Technical Report on the Refugio Gold Project – Maricunga District - Chile
	1-1

Compañia Minera Maricunga

1.0	SUMMARY			1-8
2.0	INTRODUCTION AND TERMS OF REFERENCE			2-12
	2.1	INTRODUCTION		2-12
	2.2	TERMS OF REFERENCE		2-12
	2.3	SCOPE OF WORK		2-13
	2.4	REPORT BASIS		2-13
	2.5	SERVICES BY THIRD PARTIES		2-14
	2.6	STUDY PARTICIPANTS		2-14
3.0	DISCLAIMER			3-15
4.0	PROPERTY DESCRIPTION AND LOCATION			4-16
	4.1	CLAIM STATUS		4-17
	4.2	SURFACE RIGHTS		4-17
	4.3	WATER RIGHTS		4-17
5.0	ACCESS, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE ANDPHYSIOGRAPHY			5-19
	5.1	ACCESSIBILITY		5-19
	5.2	CLIMATE AND PHYSIOGRAPHY		5-19
	5.3	LOCAL RESOURCES AND INFRASTRUCTURE		5-19
6.0	HISTORY			6-21
7.0	GEOLOGICAL SETTING			7-23
	7.1	ALTERATION		7-24
		7.1.1	Verde	7-24
		7.1.2	Pancho	7-24
	7.2	STRUCTURE		7-24
8.0	DEPOSIT TYPES			8-26
	8.1	VERDE		8-26
	8.2	PANCHO		8-29
9.0	MINERALIZATION			9-30
	9.1	VERDE		9-30

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-2 

Compañia Minera Maricunga

	9.2	PANCHO		9-30
10.0	EXPLORATION			10-31
11.0	DRILLING			11-34
	11.1	RIG SETUP AND SURVEY		11-34
	11.2	TOPOGRAPHIC BASE AND SURVEY AUDIT		11-35
	11.3	CORE HANDLING, STORAGE AND SECURITY		11-35
12.0	SAMPLE METHOD AND APPROACH			12-36
	12.1	RC DRILLING		12-36
	12.2	CORE DRILLING		12-37
	12.3	LOGGING		12-37
	12.4	GEOTECHNICAL CORE LOGGING		12-38
13.0	SAMPLE PREPARATION, ANALYSIS AND SECURITY			13-39
	13.1	SAMPLE PREPARATION		13-39
	13.2	ANALYSIS		13-40
	13.3	QUALITY CONTROL / QUALITY ASSURANCE		13-41
		13.3.1	Table of Failures	13-42
		13.3.2	Analysis of QA/QC Data	13-43
		13.3.3	Referee Laboratory	13-43
14.0	DATA VERIFICATION			14-44
	14.1	GENERAL		14-44
	14.2	PRE-2002 DATA VERIFICATION		14-44
		14.2.1	Collar Data	14-44
		14.2.2	Survey Data	14-45
		14.2.3	Lithology, Alteration, Oxidation, Vein and Mineralization Data:	14-45
		14.2.4	Analytical Data:	14-45
	14.3	2002-2003 DATA VERIFICATION		14-45
		14.3.1	Collar Data	14-45
		14.3.2	Survey Data	14-45
		14.3.3	Lithology, Alteration, Oxidation, Vein and Mineralization Data:	14-46
		14.3.4	Analytical Data:	14-46
		14.3.5	Geotechnical Data:	14-46

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-3 

Compañia Minera Maricunga

	14.4	DATA ENTRY, STORAGE, MANAGEMENT AND SECURITY		14-47
		14.4.1	Data Entry	14-47
		14.4.2	Data Storage	14-48
		14.4.3	Data Management and Security	14-48
15.0	ADJACENT PROPERTIES			15-49
16.0	MINERAL PROCESSING AND METALLURGICAL TESTING			16-50
	16.1	CYANIDE SOLUBLE ASSAYS		16-50
	16.2	DIRECT AGITATED CYANIDATION (BOTTLE ROLL) TESTS		16-50
	16.3	COLUMN LEACH TESTS		16-52
	16.4	BOND WORK INDEX		16-56
	16.5	SPECIFIC GRAVITY		16-57
	16.6	ACID ROCK DRAINAGE STUDY		16-57
17.0	MINERAL RESOURCE AND MINERAL RESERVE ESTIMATION			17-58
	17.1	GENERAL		17-58
	17.2	LITHOLOGICAL MODEL		17-58
	17.3	RECOVERY MODEL		17-58
	17.4	GRADE MODEL		17-60
		17.4.1	Sample Compositing	17-60
		17.4.2	Statistical and Geostatistical Analyses	17-60
		17.4.3	Model framework	17-61
		17.4.4	Gold Grade Interpolation	17-61
		17.4.5	Copper Interpolation	17-61
		17.4.6	Cyanide Soluble Copper Interpolation	17-62
		17.4.7	Model Checking and Verification	17-62
	17.5	RESOURCE CLASSIFICATION		17-64
		17.5.1	Measured Resource	17-65
		17.5.2	Indicated Resource	17-65
		17.5.3	Inferred Resource	17-65
	17.6	RESOURCE ESTIMATE		17-65
	17.7	GENERAL		17-65
	17.8	PIT OPTIMIZATION		17-66
	17.9	FINAL PIT DESIGN		17-67
	17.10	MINERAL RESOURCE AND RESERVE ESTIMATE		17-68

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-4 

Compañia Minera Maricunga

18.0	INTERPRETATION AND CONCLUSIONS		18-70
19.0	REFERENCES:		19-71
20.0	ADDITIONAL REQUIREMENTS FOR TECHNICAL REPORTS ONDEVELOPMENT PROPERTIES		20-76
	20.1	MOBILE EQUIPMENT	20-76
	20.2	PLANT INFRASTURCTURE	20-77
	20.3	LIFE-OF-MINE PLAN	20-77
	20.4	ENVIRONMENTAL CONSIDERATIONS	20-78
	20.5	CAPITAL AND OPERATING COSTS	20-78
	20.6	FINANCIAL ANALYSIS	20-79
21.0	AUTHORS CONSENT FORMS		21-80

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-5

Compañia Minera Maricunga

LIST OF TABLES

Table 6-1 Annual Gold Production of the Refugio Mine	6-22
Table  10-1  – Exploration Drill Summary	10-31
Table  16-1  Summary of Bottle Roll Sample Selection	16-51
Table  16-2  – Summary of Bottle Roll Results	16-52
Table  16-3  Summary of Column Leach Sample Selection	16-53
Table  16-4  Summary of Column Leach Test Results	16-53
Table  16-5  Comparison of Optimization versus Test Work Recovery	16-54
Table  16-6  – Specific Gravity Results	16-57
Table  17-1  Comparison of Verde Interpolation Methods	17-62
Table  17-2  Comparison of 2003 Estimate to 1998 MRDI Estimate	17-63
Table  17-3  Historical Pancho Resource Estimate Comparison	17-63
Table  17-4  Resource Classification Matrix	17-64
Table  17-5  – Summary of Pit Optimization Parameters	17-66
Table  17-6  Proven and Probable Reserves @ $350 per ounce	17-68
Table  17-7  Proven and Probable Reserves @ $325 per ounce	17-69
Table  20-1  Refugio Project Life-of-Mine Plan	20-77

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-6

Compañia Minera Maricunga

LIST OF FIGURES

Figure  4-1 – Location Map	4-16
Figure  8-1 – Geology of the Refugio Mine Area	8-28
Figure  16-1  Metallurgical Sample Locations  –  Verde Pit – 4300 Level	16-55
Figure  16-2  Metallurgical Sample Locations  –  Pancho Pit – 4300 Level	16-56
Figure  17-1  Verde East – Section 5NW	17-59

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-7

Compañia Minera Maricunga

1.0 SUMMARY The Refugio Project is an inactive gold mine located within Maricunga Gold Belt of the high Andes in northern Chile. The Maricunga Belt is estimated to host more than 40 million ounces of gold. The mine operated from 1996 to 2001, producing more than 920,000 ounces of gold from 46.0 million tonnes of ore. The mine was placed on care and maintenance in 2001, a result of a downturn in gold prices. In September 2002, in response to rising gold process, Compañia Minera Maricunga (CMM) approved an Exploration Program designed to increase the reserve base of the Refugio Project to a level sufficient to support resumption of active mining. The Exploration Program ran from September 2002 to June 2003. During this period, a total of 262 drill holes (51,478 meters) of drilling were completed. The drilling focused on increasing the confidence level of the known mineralization below the current Verde pits as well as increasing the confidence level in the mineralization at the nearby Pancho deposit, located approximately 2.0 km to the northeast. Much of the 2002 – 2003 drilling was diamond drill core, allowing geologists an opportunity to clearly delineate geological and alteration features affecting gold mineralization and recovery. The drill programs were carefully supervised, employing industry best practices and rigid quality management procedures in the collection and management of the field data. In addition, all historic data for the project was subjected to detailed verification programs where the data in the relational database was verified against original logs, survey calculation sheets and original assay certificates. The result is a duly verified database containing data of high quality and free of gross errors and omissions. This database was then used to estimate mineral resource and reserves for the project.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-8

Compañia Minera Maricunga

In addition to the analytical data, the drill program was also used to collect carefully controlled metallurgical samples from Verde and Pancho that would establish the metallurgical recovery expected for the remainder of the mine life. A total of 58 direct agitation (bottle roll) and 18 column leach test samples were collected and analyzed. The metallurgical samples were carefully selected to ensure adequate sample coverage of various grade bins, recovery classes and lithologies. The samples were also spatially restricted to the volume of rock within an optimized pit shell, ensuring the samples would be within reasonable distance of the final mining areas.

Ore and waste hardness was also tested and the specific gravity of the various rock types was confirmed through extensive field sampling.

The resource block model was updated to reflect the lithological models developed from observations taken from drill core. At the same time, a new recovery model was developed based on the visual logging of oxidation. The visual estimates were completed using a well-established scale previously used on the project.

The lithology and recovery models were used as hard boundaries in the resource modeling, limiting sample influence and preventing grades from one unit from influencing the grade in another unit.

Only data below the current topographic surface was used to complete the resource model.

Statistical and geostatistical analysis of the drill results was performed on the raw, 2.0 meter sample intervals to identify appropriate grade capping factors. Grade capping does not have a significant effect on the resource estimates, resulting in a less than 1% decrease in contained gold.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-9

Compañia Minera Maricunga

Directional variograms were modeled to identify anisotropy and search ranges that would guide grade interpolation. The variograms demonstrated excellent continuity, generally exceeding 100 meters in all directions.

Grade interpolation was limited to gold for the Verde deposit while gold; copper and cyanide soluble copper were modeled for Pancho. Ordinary kriging was selected as the primary interpolation method for gold. Both inverse distance and nearest neighbour grade interpolations were also performed to verify the model gold grades.

The resultant resource block model was manually verified to ensure that hard boundaries were honoured during grade interpolation. The model was classified as per CIM Guidelines as Measured, Indicated and Inferred based primarily on the search ranges indicated by the variograms.

Blocks classified as measured and indicated were exported to Whittle 4X for optimization. Optimization was based on appropriate mine operating costs and metallurgical recoveries derived from the field test work completed at the time of optimization. The optimum pit shells were imported into Datamine for final pit design using design parameters approved by a third party geotechnical engineering company.

The final mine designs were used to estimate the Mineral Resources and Reserves for the Project. The estimate was prepared by a qualified person as required by National Instrument 43-101.

It is Kinross Policy that for all open pit mine projects, mineral resource and reserve estimates must fall within optimized or final design pits that are based on reasonable, long-term, gold price assumptions. As final design pits have been completed for both Verde and Pancho, there are no mineral resources estimated at the gold prices noted.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-10

Compañia Minera Maricunga

The Mineral Reserves at Refugio at a gold price of $350 per ounce are:

Proven	98,320,000 tonnes @ 0.88 g/t	2,775,000ozs.
Probable	25,734,000 tonnes @ 0.80 g/t	658,000ozs.
Total	124,054,000 tonnes @ 0.86 g/t	3,433,000ozs.

The Mineral Reserves at Refugio at a gold price of $325 per ounce are:

Proven	79,494,000 tonnes @ 0.89 g/t	2,271,000ozs.
Probable	19,637,000 tonnes @ 0.79 g/t	499,000ozs.
Total	99,131,000 tonnes @ 0.87 g/t	2,770,000ozs.

The reserves are based on a detailed engineering study examining the economics of the project assuming a capital investment of $101.1 million to upgrade the existing infrastructure, allowing the plant to process 40,000 tonnes per day of Verde ore and 35,000 tonnes per day of Pancho ore. The capital investment is based on preliminary design and first-principle engineering. Capital cost estimates were prepared and verified by two, independent, North American engineering consultants with significant experience in Latin America.

The reserves noted above were used to complete a life-of-mine production schedule that in turn served as the basis for the financial analysis. The financial analysis indicates positive project economics at gold prices in excess of $325 per ounce.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

1-11

Compañia Minera Maricunga

2.0	INTRODUCTION AND TERMS OF REFERENCE
2.1	Introduction Compañia Minera Maricunga (CMM) requested Kinross Gold Corporation’s Technical Service Department prepare a Technical Report for the Refugio Project in support of the December 4, 2003 press release by Kinross Gold Corporation (Kinross) disclosing an updated resource and reserve estimate for the project. This report has been prepared in compliance with Canadian Securities Administrators National Instrument 43-101, under the direct supervision of:
		R. Cooper P.Eng. Director of Technical Services, Projects and Planning, Kinross Gold Corporation and
		W. Hanson, P.Geo, Manager of Technical Services, Kinross Gold Corporation.
	The individuals noted above are the “qualified persons” responsible for the content of the following report.
2.2	Terms of Reference Unless otherwise noted:
	•	all units of measurement in the following report are in metric measure;
	•	all costs are expressed in terms of United States dollars;
	•	all metal prices are expressed in terms of United States dollars;
	•	a foreign exchange rate of $1.0 US = $1.4 CAD and

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

2-12

Compañia Minera Maricunga

	•	a foreign exchange rate of $1.0 US = 700 Chilean peso.
2.3	Scope of Work The following Technical Report considered the following:
	•	Regional and local geology, structure, alteration and mineralization;
	•	sample collection, preparation, security and analysis;
	•	quality assurance and quality control procedures;
	•	data entry, verification, management, security and storage;
	•	block modeling, grade interpolation and resource estimation;
	•	metallurgical recovery, plant design and process upgrades;
	•	mine planning, scheduling and reserve estimation;
	•	environmental and operational permitting;
	•	operating and capital cost estimates and
	•	financial models
2.4	Report Basis This Technical Report has been based on verified historical data as well as data collected during recent field programs. All data was collected under the direct supervision of experienced field geologists. The qualified persons noted in Section 2.1.1 indirectly supervised all data collected in 2002 and 2003. Messrs. Cooper and Hanson completed several site visits and inspections while the 2002 – 2003 Exploration Program was underway, to observe field procedures first hand, contribute to the management of the exploration program and evaluate results.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

2-13

Compañia Minera Maricunga

2.5 Services by Third Parties During the course of the fieldwork it was necessary to rely on expertise supplied by third party professionals with technical expertise beyond the experience of the field personnel managing the work. Areas impacted by the work of third party consultants are clearly noted. Kinross relied inherently on the conclusions and recommendations of the following third party consultants:

B. Smee and Associates	Quality Control and Quality Assurance
Golder Associates	Geotechnical Design Parameters
McClelland Laboratories	Metallurgical Testing Program
SNC-Lavalin Engineers & Constructors Inc.	Metallurgical Program Design

2.6	Study Participants
	The authors wish to acknowledge the technical contributions of the following individuals: Joaquín Cabello, Gerente de Proyecto, Golder Associates S.A.
	David Evans P.Eng, Sr. Metallurgist, SNC-Lavalin Engineers and Constructors Inc. Barry Gillies P.Geo, Refugio Project Manager, Kinross Gold Corporation Tony Lipiec P.Eng, Manager, Process Engineering, Kinross Gold Corporation
	J. P. Londero, Senior Exploration Geologist, Compania Minera Marincunga
	Jack McParland, Vice President/Metallurgist, McClelland Laboratories Kevin Morris P.Eng, Manager, Mine Engineering, Kinross Gold Corporation
	V. Park P.Geo, QA/QC Database Manager, Compania Minera Marincunga
	Rob Raponi, P.Eng. Sr. Metallurgist, SNC-Lavalin Engineers and Constructors Inc. Ron Stewart P.Geo, Vice President, Exploration, Kinross Gold Corporation George Wahl, P.Geo, Sr. Geologist, SNC-Lavalin Engineers and Constructors Inc.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

2-14

Compañia Minera Maricunga

3.0

DISCLAIMER

This document has been prepared for Compañia Minera Maricunga by Kinross Gold Corporation’s Technical Services Department. The document summarizes the professional opinion of the authors and includes, conclusions and estimates that have been based on professional judgement and reasonable care. Said conclusions and estimates are consistent with the level of detail of this study and based on the information available at the time this report was completed. All conclusions and estimates presented are based on the assumptions and conditions outlined in this report. This report is to be issued and read in its entirety. Written or verbal excerpts from this report may not be used without the express written consent of the authors or officers of Kinross Gold Corporation.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

3-15

Compañia Minera Maricunga

4.0	PROPERTY DESCRIPTION AND LOCATION
	The Refugio Property is located in the Maricunga District of the III Region of Chile. The property is located 120 km due east of the city of Copiapó at elevations between 4,200m and 4,500m above mean sea level. Figure 4-1 is a generalized location map of the project.

Figure 4-1 – Location Map

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

4-16

Compañia Minera Maricunga

4.1	Claim Status
	All surface and mineral claims, surface rights and water rights are maintained in good standing. Chilean attorneys monitor claim status on behalf of CMM annually. No ownership issues with respect to the mineral claims hosting the project’s mineral resource and reserve estimates have been identified. The project mineral rights and claims are listed in Table 4-1 below:
4.2	Surface Rights
	In addition to the mineral claim rights, CMM also holds title to surface rights at Refugio, providing the land required for the leach pads, waste dumps, camp and other facilities. All surface rights affecting the resumption of mining activity have been reviewed. No fatal flaws have been identified that would negatively affect the project.
4.3	Water Rights
	Water extraction rights, totalling 258 L/s have been secured by CMM. Permits are in place and maintained. No issues regarding water rights have been identified that would adversely affect the project.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

4-17

Compañia Minera Maricunga

Table 4-1 Refugio Project Properties

Claim Name	Assigned Blocks
Refugio 1-112	1 through 112
Maricunga 1-40	17 through 20
Maricunga 1-40	27 through 30
Maricunga 81-120	97 through 100
Maricunga 121-160	121 through 124
Maricunga 241-280	261 through 262
Demasias on Refugio	1 through 112 Western Boundary
Buffer Zone Properties
Claim Name	Assigned Blocks
Maricunga 1-40	1 through 16
Maricunga 1-40	21 through 26
Maricunga 1-40	31 through 40
Maricunga 81-120	93 through 96
Maricunga 81-120	113 through 120
Maricunga 121-160	125 through 160
Maricunga 161-200	161 through 167
Maricunga 161-200	171 through 177
Maricunga 161-200	181 through 187
Maricunga 161-200	191 through 197
Maricunga 201-240	201
Maricunga 201-240	221 through 223
Maricunga 241-280	241 through 260
Maricunga 241-280	263 through 280
Anillo Seis 1-30	3 through 15
Anillo Seis 1-30	18 through 30
Anillo Siete 1-20	1 through 5
Anillo Siete 1-20	11 through 14
Anillo Ocho 1-20	1 through 2
Anillo Ocho 1-20	6 through 7
Anillo Ocho 1-20	11 through 12
Anillo Ocho 1-20	16
Anillo Trece 1-20	8 through 10
Anillo Trece 1-20	20
Anillo Catorce 1-20	1 through 17
Anillo Quince 1-20	1 through 2
Anillo Dieciseis 1-30	10 through 11
Amparo Unio 1-60	45 through 60
Amparo Dos 1-26	25 through 26
Amparo Seis 1-60	31 through 32
Amparo Once 1-60	5 through 30
Amparo Once 1-60	35 through 60
Amparo Doce 1-36	1 through 36
Laguna 1-11	1 through 11
Hielo 1-9	1 through 9

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

4-18

Compañia Minera Maricunga

5.0	ACCESS, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY
5.1	Accessibility
	Access to the property is via 156 km of two lane dirt road connecting with the paved highway C-35 approximately 10 km south of Copiapó. The first 96 km of the dirt road are an international, public highway. Approximately 60 km from the Refugio site, the road branches to the southeast to Argentina and the northeast to the mine site. The final 60 km is a private road. The condition of the access road is generally good.
5.2	Climate and Physiography
	The Refugio Project is located in steep, mountainous terrain with slopes up to 35%. The site is largely devoid of vegetation with the exception of the spring-fed marshes found along the valley floors. The climate is arid with an average annual precipitation of 87mm., most of which is realized as snowfall during the winter months (March through August). Generally, very little precipitation occurs during the summer months (September through February). Local wildlife is sparse.
5.3	Local Resources and Infrastructure
	The mine previously operated from 1996 to 2001 at which time operations were suspended and the mine was placed under “care and maintenance”. Residual leaching of the leach pads has continued since the suspension of active mining. As a result, the leaching and ADR facilities are in good repair as are the power generation, maintenance shops, office facilities and camp. The existing infrastructure also includes an established fresh water supply system, process water supply system, truck shop, warehousing and lay down areas, an in pit crushing and conveying system and a secondary/ tertiary crushing and screening plant. Most of the existing infrastructure requires little to no modifications or improvements other than general clean up and repair. Significant upgrades designed to increase production throughput are planned for the in-pit crushing and conveying and secondary/tertiary crushing and screening infrastructure in order to meet planned production throughput.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

5-19

Compañia Minera Maricunga

The town of Copiapó is the primary staging and support area for the mine. Copiapó features an established construction contractor (Vechiola) with road construction and site construction experience. Several, small welding and steel fabricators are also located in Copiapó. There is also a Caterpillar distributor established in Copiapó.

Chile features a strong mining culture with well established support centers in both Santiago and Antafagasto. Both centers are within reasonable distance of the project. Most of the major equipment supply and support will originate from these two major centers.

Manpower is expected to be attracted from throughout Chile with the majority of the employees residing in Copiapó or La Serena.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

5-20

Compañia Minera Maricunga

6.0

HISTORY

David Thomson and Mario Hernandez discovered gold mineralization at Refugio in 1984. Shortly after their discovery, Hernandez, Thomson, and three other partners acquired the existing claims at Refugio for Compania Minera Refugio (CMR). CMR completed geologic mapping and geochemical sampling, identifying anomalous gold values in three areas: 1) Cerro Verde, 2) Cerro Pancho, and 3) Guanaco. In 1985, Anglo American Chile Limitada (Anglo) optioned the property from CMR. Anglo explored the property for three years, returning the claims to CMR in 1988. In 1989, CMR signed a letter of intent to explore the Refugio property with Bema Gold Corporation (Bema). Bema commenced exploration fieldwork in October 1989 and from 1989 to 1991, completed 51,765 meters of drilling at Verde with an additional 5,088 meters at Pancho. Bema also commissioned Mineral Resources Development Inc (MRDI) to complete a feasibility study on the project, which indicated positive project economics. In January 1993, Bema exercised its option rights, obtaining a 50% interest in the Refugio properties. At the same time, CMR sold its remaining 50% interest to Amax Gold Inc. (Amax). Amax (operator) and Bema formed Compania Minera Maricunga (CMM), a 50/50 joint venture to develop and operate the Refugio project. From 1993 through 1997, CMM continued developing the project, beginning commercial production in 1996. The mine operated from 1996 to 2001, producing more than 920,000 ounces of gold from 46.0 million tonnes of ore. Table 6-1 summarizes the production history of the mine.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

6-21

Compañia Minera Maricunga

Table 6-1 Annual Gold Production of the Refugio Mine

Year	Mined tonnes (x 1,000)	Grade mined (Au g/t)	Gold on pad (ounces)	Gold production (ounces)	Annual recovery (%)	Cumulative recovery (%)
1996  1997  1998  1999  2000  2001  2002  2003  TOTAL	7,617  7,789  8,207  8,936  8,801  4,643  -  -  45,993	1.028  1.021  0.928  0.940  0.938  0.946  -  -  0.966	251,749  255,681  244,863  270,061  265,415  141,215  -  -  1,428,984	101,276  147,085  161,046  179,465  169,832  133,947  26,094  5,000  923,745	40.2  57.5  65.8  66.5  64.0  NA  NA  NA	40.2               48.9               54.4               57.6               58.9               62.5               64.3               64.6               64.6

In 1998, Kinross Gold Corporation (Kinross) acquired Amax’s 50% interest through a merger agreement.

Mining operations at Refugio were suspended during the second quarter of 2001, a result of depressed gold prices. Improving gold prices during 2002 prompted CMM to evaluate options for re-opening the mine. Scoping studies quantifying the conditions necessary to re-open the mine completed. The studies indicated that the project economics required an increase in the existing reserve base. As a result, an exploration program was developed to evaluate the reserve potential at depth at the Verde deposits and the inferred resource at the nearby Pancho deposit, located approximately 2.0 km northwest of the Verde pit.

A 35,000 meter drill program was approved in September 2002 with the objective of adding 20 million tonnes of measured and indicated resource at both Pancho and Verde. This mineralization had to be recoverable through open pit mining methods. Follow-up work to further delineate the mineralization, provide representative metallurgical samples and detailed geotechnical information would be contingent on the initial results of the drilling.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

6-22

Compañia Minera Maricunga

7.0

GEOLOGICAL SETTING

The Verde and Pancho gold deposits at Refugio occur in the Maricunga Gold Belt of the high Andes in northern Chile. Since 1980, a total of 40 million ounces of gold have been defined in the belt, (Muntean and Einaudi, 2000). Basement rocks in the Refugio area are Paleozoic to early Tertiary age. The oldest rocks are the Late Pennsylvanian to Triassic aged rhyolite ignimbrites and breccias of the Pantanosa Formation. This unit has been uplifted along a northerly trending, westerly dipping reverse fault. The Pantanosa Formation is faulted over a package of interlayered redbeds and greenstones of the Monardes and Agua Helada Formations. These formations are Late Jurassic to Early Cretaceous in age. They comprise an east dipping (55°) sequence of strata with a thickness of 900 m. A 200 meter thick sequence of andesitic volcaniclastic sedimentary rocks overlie the redbeds and greenstones. These strata are correlated with Late Cretaceous to Early Tertiary rocks of the Quebrada Seca, Quebrada Paipote, and Las Pircas Formations. Gold mineralization at Refugio is hosted in the Refugio volcanic-intrusive complex of Early Miocene age. These rocks are largely of intermediate composition. Radiometric age dating indicates these rocks are between 22 to 24 Ma in age. (Muntean 1998). The Refugio volcanic-intrusive complex is exposed over an area of 12 km2 and consists of andesitic to dacitic domes, flows, and breccias that are intruded by subvolcanic porphyries and breccias (Muntean 1998). Distinguishing between volcanic and intrusive rocks in the complex is difficult as there are only minor differences between the various units.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

7-23

Compañia Minera Maricunga

7.1	Alteration
7.1.1	Verde
	Alteration assemblages observed at Verde and Pancho are generally supportive of porphyry style mineralization but the intensity of the alteration fabric tends to be weak. Potassic alteration has been observed at Verde but generally tends to be rare. Silicification is local and patchy. Propylitic alteration, although variable on small scale, appears ubiquitous on a mine scale and in global terms does not change laterally or vertically. For that reason, primary alteration (including silicification) zonation has not been mapped.
	Supergene alteration, which directly affects gold recovery, occurs deeper within fracture and fault zones. Sericite and chlorite are replaced by clay minerals, magnetite by hematite and pyrite by jarosite. The oxidation and the leaching by meteoric waters has penetrated to variable depths within the deposits depending on the fracture intensity and faulting. The supergene alteration is accompanied by the deposition of limonite, manganese oxides, clay, sericite, jarosite along with gypsum in the argillic altered zones.
7.1.2	Pancho
	Gold mineralization at Pancho is associated with a central zone of potassic alteration, which is manifested by replacement of mafic minerals, by fine grained, secondary biotite and magnetite. Partial replacement of plagioclase by k-feldspar can also be observed. Muntean (1998) has documented a more restricted core to this potassic zone consisting of magnetite-k feldspar-oligoclase replacement which grades outward into the more widespread secondary biotite zone. Similar to Verde, much of the potassic alteration is obliterated or obscured by a later chlorite overprint.
	The intermediate to upper parts of the system are dominated by pyrite-albite-clay- sericite alteration, primarily in the volcanics but also overprinting the upper parts of the potassic zone. This alteration is late and probably has a large supergene component.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

7-24

Compañia Minera Maricunga

In the uppermost parts of the system analysis indicates the presence of hypogene alunite, dickite and pyrophyllite, characteristic of epithermal, high sulfidation alteration. This hypogene assemblage occurs together with strong supergene alunite, kaolinite and other clays. The presence of this high level alteration within 100-200 m of the potassic zone suggests strong telescoping of the system.

7.2

Structure Most of the structural trends affecting the Verde and Pancho deposits are related to fracture systems rather than fault zones. The dominant trends are north-northeast and north-northwest fracture systems that can be readily identified at Verde. At Pancho, the dominant structural trend is northwesterly, reflected not only in mineralized structures but also in late, post mineral structures. One of the main structural features influencing the Pancho deposit is Falla Guatita fault zone. Field mapping suggests that there may be significant vertical displacement on this structure. The distinctly higher copper grades and presence of potassic alteration in the main zone suggests that the central portion of the Pancho deposit may have been down dropped. Another major fault affecting the Pancho deposit is the Falla Moreno, which is interpreted to have significant, north side down offset. This structure trends roughly east west and forms an approximate northern boundary for the mineralization at Pancho.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

7-25

Compañia Minera Maricunga

8.0	DEPOSIT TYPES
8.1	Verde
	Gold mineralization at Refugio has been interpreted to be porphyry style gold systems. At Verde West, gold mineralization is centered about an elliptical porphyry plug measuring 175 by 100 m and oriented at N30°W. At Verde East, the porphyry plug measures 130 by 80 m and is oriented at N35°E. The porphyries occur within a sequence of intermediate tuffs, porphyries and breccias that are the host rocks to the gold mineralization. Lithological interpretation at Verde has identified six major lithologic units. These are:
	•	Post mineral intrusives (barren),
	•	Mineralized post mineral intrusives,
	•	Verde breccia
	•	Dacite porphyry
	•	Dacite tuffs
	•	Laguna tuff (barren)
	The most favourable ore hosts are the Verde breccia and dacite porphyry units. The dacite porphyry is a volcanic to hypabyssal intrusive rock with 20% to 40% plagioclase phenocrysts in a fine-grained matrix. It contains phenocrysts of biotite, hornblende, and sparse quartz. The unit is the best host for stockwork veining, with some portions containing up to 20% quartz-magnetite ± pyrite veinlets. The Verde Breccia consists of intrusive breccia and/or volcanic tuff breccia. It’s geometry suggests an intrusive origin. The unit consists of breccia with angular to rounded clasts, generally matrix supported. The breccia is generally green to greenish gray, which is one of the diagnostic features of the unit. The color of the rock is largely due to chlorite and occassionally epidote. The fragments range from 2

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

8-26

Compañia Minera Maricunga

mm to greater than 2 meters in size. Locally the unit is mostly matrix-sized material with only sparse clasts. Clasts in the unit are generally monolithic volcanic fragments, which are often porphyritic with white plagioclase laths up to 5 mm in length. The unoxidized rock typically contains 0.5 to 1% pyrite. Quartz-magnetite veinlets are common in the mineralized portions of the unit. In places, the breccia is cut by fine-grained matrix material that forms clastic-like dikes. These are generally > 2 cm in width. They may have been formed as “fluidized” material injected into fractures at the time of the formation of the unit.

Figure 8-1 is a geological map of the Refugio Area (after Muntean, 2000).

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

8-27

Compañia Minera Maricunga

Figure 8-1 –Geology of the Refugio Mine Area

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

8-28

Compañia Minera Maricunga

8.2 Pancho Mineralization at Pancho is concentrated within a sub-horizontal volcanic breccia (P_BX) unit. Underlying the volcanic breccia is a large, laterally extensive, diorite porphyry, which outcrops half way down the Pancho west slope. This porphyry underlies the entire Pancho area. The diorite appears to have been a semi-homogenous fracturable host for quartz veinlet gold bearing mineralization. Lithological mapping has subdivided the porphyry into three distinct units; Diorite Porphyry (P_DIO), Micro-Diorite (P_DIO_A) and Biotite Porphyry (P-DIO_B), based largely on stratigraphic position within the deposit. Although not exposed on surface, a post mineral intrusive (P_DIO_B) similar to Verde has been interpreted at depth. It appears to be a fresh equivalent of the diorite porphyry, without quartz veinlets or mineralization. It does not form a well defined, subvertical column as at Verde but occurs as several dyke like fingers. Mineralization appears to be confined beneath a thick bedded to massive pyroclastic volcanic (P_VOLC) unit, which, due to poor fracturing characteristics appears to have acted as an aquatard during mineral emplacement.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

8-29

Compañia Minera Maricunga

9.0	MINERALIZATION
9.1	Verde
	Gold mineralization at Verde is interpreted to be the result of the fracturing and concentration of fluids in the carapace of an intrusive plug or stock. Gold is closely associated with quartz, magnetite, calcite, and garnet stockworks. Approximately 80% of the stockwork veins are generally dark grey in color, finely banded, with magnetite. The remaining 20% are principally white quartz veins. Gold mineralization is postulated to have resulted from at least 2 phases of mineralization, the first is a lower grade phase associated with copper and probably the porphyry emplacement event. The second phase is a higher-grade gold only event possibly associated with the structurally emplaced veinlet swarms and northwest trending sheeted veinlet zones more evident in Verde East.
9.2	Pancho
	Gold mineralization at Pancho is characterized as porphyry hosted, stockwork and sheeted veins. The veins are subvertical and have a strong, preferred north-westerly strike. The northwest structural control is evident not only at outcrop scale but is also reflected in the northwest alignment of intrusives and the three centers of mineralization in the district, Verde, Pancho and Guanaco. Several vein types are associated with mineralization at Pancho but their relative distribution was not differentiated in this program. Gold occurs in banded, vitreous, quartz veinlets, in quartz magnetite veinlets and in what Muntean (1988) describes as early, quartz- magnetite-chalcopyrite “A” veins after the terminology of Gustafson and Hunt (1975). Veinlet frequency is low, usually 1-2 per drill meter.
	The porphyritic diorite intrusives and intrusive breccias are the main hosts for mineralized veins. Mineralization is also hosted by the volcanic rock but grades drop rapidly away from the intrusive contact. Mineralization is interpreted to be confined beneath a thick bedded to massive pyroclastic unit, which, due to poor fracturing characteristics would have acted as an aquatard.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

9-30

Compañia Minera Maricunga

10.0

EXPLORATION

Exploration of the Verde and Pancho deposits has been ongoing since 1984. A total of 667 holes (103,392 m) of drilling has been completed on the Verde deposit with an additional 147 holes (30,240 m) completed at Pancho. The drilling has resulted in a drill spacing of approximately 50 x 50 meters at Verde and 75 x 75 meters at Pancho. Table 10-1 summarizes the various drill campaigns at Verde and Pancho.

Table 10-1 – Exploration Drill Summary

		Verde				Pancho
Company	Year	RC (#)	RC  (m)	Core (#)	Core  (m)	RC (#)	RC  (m)	Core (#)	Core (m)
Bema  Bema  Bema	1989  1990  1991	45  231  -	5,060  46,705  -	-  31  6	-  4,083  1,090	-  -  24	-  -  5,088	-  -  -	-  -  -
SUBTOTAL		276	51,765	37	5,173	24	5,088	-	-
CMM  CMM  CMM  CMM	1993  1994  1997  1998	176  -  -  -	5,060  -  -  -	-  6  -  18	-  4,083  -  6,689	-  -  15  -	-  -  4,296  -	-  -  -  -	-  -  -  -
SUBTOTAL		176	5,060	24	10,772	15	4,296	-	-
CMM	2003	20	3,154	134	27,468	39	6,710	69	14,146
TOTAL		472	59,979	195	43,413	78	16,094	69	14,146

Figures 10-1 and 10-2 are plan maps of Verde and Pancho showing the completed drill holes as at December 1, 2003.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

10-31

Compañia Minera Maricunga

Figure 10-1 Verde Drilling Coverage as at June 30, 2003

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

10-32

Compañia Minera Maricunga

Figure 10-2 Pancho Drill Coverage as at June 30, 2003

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

10-33

Compañia Minera Maricunga

11.0	DRILLING
	Historically, most of the drilling at Refugio consisted of reverse circulation drilling. The destructive nature of this drill method made identification of lithology, structure and alteration difficult. The 2002 - 2003 drilling consisted primarily of diamond drill core, providing site geologists with an opportunity to refine the geology model of the deposit. The preferred orientation of drilling for the Verde deposits is 060 - 240° azimuth. At Pancho, the preferred orientation for the drilling grid of 038 – 218° was established based on the northwest strike of veinlet swarms observed in drill roads.
11.1	Rig Setup and Survey All proposed drill locations were laid out in plan and on section. The collar coordinates were provided to the mine survey crews who laid out the drill pad locations in the field using total station theodolites. Pad locations were verified prior to construction to ensure access and safety. After pad construction, the mine survey crews established the collar location and marked it in the field. They also established the front sight and back sights necessary to provide the drill direction. After the drill was moved onto the setup and prior to the start of drill, geological staff verified the drill alignment and inclination using a Brunton compass. The survey crew later verified alignment and inclination when the drill hole was in progress. Downhole inclinometry was completed at the end of each hole. Gyroscopic azimuth and inclination readings were taken every 10 meters down the hole to within ten meters of hole bottom and every 50 meters back up the hole as a double check.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

11-34

Compañia Minera Maricunga

11.2	Topographic Base and Survey Audit
	All field surveys were tied into the established mine grid. Survey data was incorporated into current as-built plans that were updated and maintained in AutoCAD by the survey crews. Guillermo Contreras and Sons Limitada (Santiago), licensed Chilean surveyors completed a survey audit that verified an approximate 10% of the drill collars using a differential GPS survey system. No significant errors were noted.
11.3	Core Handling, Storage and Security
	Drill core was placed in labelled boxes, fitted with lids and transported to the site logging facility by pick-up. All core logging was completed at an open air logging facility located behind the administrative building. Only CMM personnel worked with the drill core.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

11-35

Compañia Minera Maricunga

12.0

SAMPLE METHOD AND APPROACH

12.1

RC Drilling CMM provided all of the technical support for the sampling, geologic logging, and drill supervision. Rig geologists and samplers were responsible for the quality/accuracy of each sample. Geologists and samplers typically had up to 15 years experience sampling. The 2003 drill program adopted a 2.0 meter standard sample length for all samples. When drilling dry, reverse circulation drill cuttings were directed from the cyclone to a Gilson splitter where the sample was reduced to a 6-8 kg sample. Typically, this required a 1/ 8 split of the drill cuttings. The hole and cyclone were blown clean after each rod change. Two samples, an original and a duplicate, were produced for each 2.0 meter interval. The samples were weighed, and the weight, number of splits and hole diameter were recorded. The samples were bagged and tagged and prepared for transport to the on-site sample preparation facility. The splitter was air cleaned after each sample. When sampling wet, a rotating wet sampler replaced the Gilson splitter below the cyclone. Usually ground water flow was not sufficient and make up water was necessary to provide uniform flow to the splitter. As with the dry sample collection, two samples (6 – 8 kg.) were collected in perforated sample bags at the exit port of the wet splitter. These samples were not weighed immediately but were left to drain for a 24 hour period prior to transport to the sample preparation facility. Typical chip trays were produced for logging and future reference. In addition, a 100 g, unwashed sample (reference sample) was collected for each 2.0 meter interval. All chip trays and reference samples were placed in permanent storage at site.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

12-36

Compañia Minera Maricunga

12.2	Core Drilling
	Drill core was received and laid out for logging and sampling at the core logging facility. The first sample of each core hole was marked from the start of core recovery to the end of the first even meter. From that point, samples were laid out every 2.0 meters downhole. Sample numbers were marked on the core box edges, accounting for number gaps for QA/QC sample insertion. Logged and marked boxes of whole core were transferred to the ALS Chemex sample preparation facility where chain of custody was transferred from CMM to ALS Chemex.
	CMM received the sawn core, in the original labelled boxes, from ALS Chemex after splitting. The half-cores were stored shipping containers at the Refugio site. In all, ten containers were used to store the half-core library.
	Core sample rejects are bagged in plastic, labelled and received from the ALS Chemex sample preparation facility. These bags are placed in cubic meter wooden bins. The bins are labelled according to hole number and sample range. The bins are stored beside the core containers and are locked in a covered storage compound.
12.3	Logging
	Experienced geologists logged the reverse circulation chips and drill core at Refugio. Reverse circulation chips were typically logged at the drill while core was typically logged at the core logging facility. All logging utilized standardized logging forms and a geological legend developed for the Verde and Pancho deposits. The legend has evolved from historic observation and is consistent with both the regional and local geology. The legend and logging records lithology, alteration, structure, mineralization and oxidation (of pyrite) for each two meter sample interval, reducing the geological descriptions to numeric or alphanumeric codes. Unique features not accounted for in the legend are noted in written comments. The legend stresses oxidation conditions of pyrite, veinlet mineralization, location and amount of magnetite and occurrence of typical porphyry alteration minerals.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

12-37

Compañia Minera Maricunga

Golder Associates Santiago (Golder) identified all holes requiring detailed geotechnical logging that was completed by Golder trained staff. CMM geologists typically recorded both Rock Quality Designation (RQD) and recovery data for each hole. Recoveries averaged better than 95% during the 2003 program. During the 2002 - 2003 spot check logging and cross checks with pit mapping by senior expatriate geologists found the logging to be of good quality.

12.4

Geotechnical Core Logging Golder managed the geotechnical program. Francisco Carrasco, (Project Manager) and Joaquin Cabello (Assistant Manager) supervised the filed programs and provided the initial training. The program included detailed geotechnical logging of selected drill holes, uniaxial compressive strength (UCS) testing of selected cores, completion of a seven (7) hole oriented core program and cell mapping of existing pit walls and road cuts. Golder also managed a hydrology study including installation of peizometer wells for monitoring water levels. Golder provided all initial training and completed frequent site visits to maintain quality control of the work and upgrade training of the filed staff performing the work.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

12-38

Compañia Minera Maricunga

13.0	SAMPLE PREPARATION, ANALYSIS AND SECURITY
13.1	Sample Preparation
	All sample preparation, including core splitting (sawing) was performed and supervised by ALS Chemex personnel. ALS Chemex established, equipped and staffed a portable sample preparation facility at the Refugio mine for the duration of the program. The sample preparation facility included two rock saws, three drying ovens, two Rhino jaw crushers and a Jones riffle splitter.
	After logging and sample mark up, the core boxes were sent to the sample preparation facility. ALS Chemex personnel physically sawed the core. The core splitter randomly selected the split line. Core splitters were trained to maintain the same cut line on contiguous pieces of core wherever possible. After splitting, one half of the core was placed in sample trays along with the sample tag. These were placed in the drying ovens for drying. The other half of the core was fitted back into the core box and returned to CMM for placement in permanent storage.
	All samples were dried for 4 – 6 hours at 70°Celsius. For wet reverse circulation samples, the dry time was increased to 24 hours. After drying, sample weight was recorded and the sample was single pass crushed to 90% passing 2 mm using the Rhino Jaw crushers. The crushers were cleaned with air after each sample and cleaned with barren quartz after every five (5) samples. Crush size was monitored after every 20 samples to ensure that the final product met the crush size specification.
	After crushing, the entire sample was passed through a Jones riffle splitter, reducing the sample size down to a nominal 1.0 kg size. A preparation duplicate was collected every forty (40) samples. The duplicate was distinguished by an “A and B” designation and treated as per the standard sample procedure. The coarse rejects were returned to the original sample bag, labelled and returned to CMM for storage.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

13-39

Compañia Minera Maricunga

	The 1.0 kg samples were sealed in plastic bags, labelled, packed in crates and transported to the primary analytical facility (ALS Chemex – La Serena) for analysis.
13.2	Analysis The prepared samples were received at the ALS Chemex’s facility in La Serena (the primary analytical lab for the duration of the program) where analyses for Au, Ag, Cu and cyanide soluble Au and Cu analyses were performed. On receipt of the samples the work order numbers and sample numbers were recorded and the samples were arranged for analysis based on the individual sample numbers. The entire 1.0 kg sample was pulverized to 95% passing -150 mesh using an LM-2 pulverizer. The pulverizer was cleaned with compressed air after each sample. Sieve analyses were performed after every 20th sample to ensure the final product met the stated specification of 95% passing –150 mesh. After pulverization, the sample was reduced to a 300 gram sample and 700 gram reject. The reject placed in labelled plastic bags and returned to CMM for storage. The 300 gram sample is homogenized. After homogenization:
	•	a 50 gram aliquot is selected for fire assay with atomic absorption finish for gold,
	•	a second aliquot is selected for total digestion atomic absorption analyses for silver and copper,
	•	a 20 gram aliquot is selected for cold cyanide soluble gold analysis, leached for 4 hours with 60 ml of cyanide solution (0.1% NaOH & 0.5% NaCN).
	•	A 1 gram aliquot is selected for cold cyanide soluble copper analysis, leached for 2 hours with 20 ml of cyanide solution.
	Dr. B. Smee (Smee and Associates Consulting Ltd.) completed operational audits of the La Serena lab for the 2002-2003 program. The operational audits were

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

13-40

Compañia Minera Maricunga

performed measuring compliance with analytical best practices as well as to NI-43-101 requirements with respect to quality control and quality assurance. Dr Smee did not note any significant problems with this facility, concluding that ALS Chemex’s lab and procedures met the highest industry standards. Initially, ALS Chemex assigned work order numbers at the mine site but in March 2003 a Lab Information Management System (LIMS) was adopted and work order numbers were generated from a centralized ALS Chemex location. Results containing Au, Ag, Cu or CN Au, CN Cu were sent via digital mail to the Project and Office Managers. Attachments included 1 file per hole, identified by work order number and hole number. The file included internal ALS Chemex blanks and duplicates, while results of ALS Chemex standards were reported monthly in a separate file. ALS Chemex was also requested to send from a list chosen by CMM 3 batches of pulps (1/40 samples) to Assayers Canada in Vancouver. Dr. Smee audited the Assayers laboratory in February 2003 and found them acceptable to qualify as a referee lab for the project.

13.3

Quality Control / Quality Assurance The QA/QC program was designed and monitored by Dr. B. Smee. The program included the insertion of one blank, one standard, and one duplicate (randomly selected) for each batch of 40 samples. Standards were not available for insertion for the first 3 weeks of the program. The standards were designed by Dr. Smee and provided by Rocklabs, New Zealand. Dr. Smee selected the standards to best match Refugio rock types and grades. Standards were designed to monitor gold grade only. The standards were not blind to the analytical lab.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

13-41

Compañia Minera Maricunga

Duplicate samples were randomly selected within the sample batches. Reverse circulation duplicates were taken as an opposite split post cyclone and given a sequential sample number. RC duplicates were not taken below the water table. As the duplicate was generated in the field, no ALS Chemex personnel could identify these samples as duplicates. For core, two duplicates of one interval were taken. The first (d1) was a second split of post crusher product. The second (d2) was the actual other half of sawn core. The intent was to see the variance according to crushing procedure (d1) and also the variance according to natural geological variance + crushing procedure. This would also allude to any bias in core sawing activity. In the core duplicate samples, obviously ALS Chemex sample preparation personnel knew which samples were being duplicated, while ALS Chemex analytical personnel did not. ALS Chemex also inserted their own blanks, standards and duplicate samples for each sample batch as part of the labs own internal quality management program.

13.3.1

Table of Failures The Senior Geologist extracted the QA/QC sample results from the digital result files based on the sample numbers. If a standard assay fell outside 3 standard deviations of the standard value, the entire batch of 40 samples was rerun. If 2 consecutive standard assays fell within 2 standard deviations of the standard value, a batch rerun was requested. If a blank exceeded 0.05 g/t au, a batch rerun was requested. No reruns were determined from duplicate data. Monitoring of the QA/QC data required a total of 17 batch reruns (649 samples) based on standards failures. This represented approximately 2% of the total analyses run for the project. Inspection of batch failure results indicates minimal bias with both gold and copper relative to the repeat batch assays. At the end of the project, an additional nine QA/QC batch assays (347 samples) were selected based on standards failures. Three additional batches plus 11 individual samples (127 total) were found to have CNAu at least 0.10 g/t au greater than the total gold assay. These were also rerun for both total and cyanide.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

13-42

Compañia Minera Maricunga

13.3.2	Analysis of QA/QC Data
	All standard, blank and duplicate results were sent to Dr. Smee for analysis. Duplicate results were analyzed for precision and plotted on Thompson Howard graphs. Dr. Smee found the error rate and duplicate variance normal for porphyry gold deposits. In his final report, (July 20,2003) he concluded the gold assay database to be of commercial quality but raised concerns about eight of the 650 ALS Chemex digital files. All eight files, containing failures, had already been reanalyzed, were determined to have had CMM field labelling errors, or were duplicated assay sets.
13.3.3	Referee Laboratory
	Assayers Canada (Assayers), Vancouver, was selected as the referee laboratory for the program. One pulp per ALS Chemex batch of 40 was selected at random, including the CMM standards, blanks and duplicates. Additionally one of the 3 CMM standards was inserted in this sample stream to check Assayers at a frequency of 1 per 40. Three batches of checks (631 samples in total) were sent to Assayers over the duration of the program. Results were forwarded to Dr. Smee who concluded that the two labs essentially agreed on gold and cyanide soluble gold analysis within normal precision limits.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

13-43

Compañia Minera Maricunga

14.0	DATA VERIFICATION
14.1	General
	Kinross Technical Services verified the Initial Database (IDB) to copies provided by Bema Gold Corporation and MRDI. The verified IDB was forwarded to the CMM field team to serve as the starting point for the project. The IDB was backed up to CD-R format and stored.
	Original drill logs, assay certificates, survey worksheets and other original data were located and used to verify the data contained in the IDB. Verification of the historical data included:
	•	Collar location,
	•	Collar azimuth and dip,
	•	Downhole azimuth and dip,
	•	Total gold and copper analyses,
	•	Cyanide soluble gold and copper analyses,
	•	Primary lithology code,
	•	Primary alteration code,
	•	Primary oxidation code

	Assay certificates, and original logs were found for approximately 80% of the historical data.
14.2	Pre-2002 Data Verification
14.2.1	Collar Data
	Verification of the historic collar data identified one error of note. Collar coordinates of drill holes completed between 1986-87 did not match the coordinates recorded on original survey sheets. This error was also corrected.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

14-44

Compañia Minera Maricunga

14.2.2	Survey Data
	Much of the historical downhole survey data could not be located. The limited downhole survey data that was located was used to verify the data in the IDB. No significant errors were noted.
14.2.3	Lithology, Alteration, Oxidation, Vein and Mineralization Data:
	Ten percent of all lithology, alteration, oxidation, vein description and mineralization data for the 1997 – 1998 drill campaign was verified as this data was collected using field practices and logging formats consistent with the 2002-2003 effort.
14.2.4	Analytical Data:
	Ten percent of the historical analytical results in the IDB were verified against original assay certificates. Rare errors were detected during the verification process but these were almost exclusively related to the recording of results below the detection limit for successive programs.
14.3	2002-2003 Data Verification
14.3.1	Collar Data
	Prior to entering collar data into the database, collar locations were visually checked for obvious errors such as transposed northings and eastings. Once entered, collar coordinates and hole lengths were extracted and printed for direct comparison to the original survey certificates. The collar coordinates were compared against the signed surveyors’ certificate and the hole length were checked against the geologists’ logs. Three independent checks of this information yielded rare errors. Collar information was also digitally checked against the surveyors’ final compilation file; this check also yielded rare decimal place errors. Erroneous data was corrected and reloaded. The corrected data was subjected to two additional independent checks.
14.3.2	Survey Data
	Prior to entering the downhole survey data into the database it was visually scanned for obvious errors. After entry, the data was extracted and printed for manual verification relative to the original downhole survey data contained in the files and

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

14-45

Compañia Minera Maricunga

	provided by Comprobe. Data verification of 10% of the downhole survey data was completed for the survey data. Two independent checks of this data detected no errors. Visual inspections of the plotted hole traces also failed to identify any obvious errors or miss-plots.
14.3.3	Lithology, Alteration, Oxidation, Vein and Mineralization Data:
	Lithology, Alteration, Oxidation, Vein Description and Mineralization data collected during the 2002-2003 field season was verified against the original logs. A 10% check of the database yielded an unacceptable amount of errors and led to a 100% check. The 100% check yielded an error rate of 3.5%, with most of the errors related to the recording of alteration. These errors were corrected
14.3.4	Analytical Data:
	A random 10% of the analytical data was verified against the original assay certificates. Only rare errors were detected, usually pertaining to how results below detection were entered in the database. Erroneous data was corrected and re- entered.
14.3.5	Geotechnical Data:
	Ten percent of the geotechnical data collected during the 2002-2003 field program was verified against the original geotechnical logs yielding a significant error rate. As a result, a 100% verification was initiated. Significant error rates were noted in the fractures-per-meter field. Error rates in the structural measurement and RQD fields were related to format errors rather than data entry errors. These errors were corrected.
	The QA/QC Database Manager has certified that the database as transferred to Kinross Technical Services on CD-R media is free of gross errors and omissions and is suitable for estimating resource and reserve estimates.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

14-46

Compañia Minera Maricunga

14.4	Data Entry, Storage, Management and Security The QA/QC Database Manager was responsible for updating and managing the Project Database (PDB). Only the QA/QC Database Manger could access, update or otherwise modify the PDB. Regular backups were made to CD-R disks over the duration of the program. The final PDB was transferred to Kinross Technical Services to be used in resource modeling.
14.4.1	Data Entry
14.4.1.1	Collar Coordinates The mine survey department entered collar coordinates, drill azimuth and inclination into spreadsheets that were provided to the QA/QC Database Manager. The QA/QC Database Manager checked the spreadsheet data against original survey field notes prior to importing the data into the PDB. Cross sections and drill hole composite plans were plotted using Gemcom software and were inspected by the Senior Geologist for collar and directional errors. Final hole depths were verified against the drill log hole depth.
14.4.1.2	Downhole Survey Data Comprobe provided the downhole survey data in spreadsheet format to CMM’s Office Manager who checked the data for obvious errors. Once satisfied that the downhole survey data was correct, the Office Manager provided the spreadsheet files to the QA/QC Database Manager for import into the PDB.
14.4.1.3	Lithology, Alteration, Mineralization and Oxidation The Office Manager and Data Entry Clerks entered data from the field lithology logs into spreadsheets, one for each hole. A double-blind entry system was not used in recording this data. A 100% manual check of this data was performed at the end of the drill program.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

14-47

Compañia Minera Maricunga

14.4.1.4	Assay Data Assay results were received as digital files, identified by work order number and hole number. The assay files included individual sample numbers allowing the samples to be matched to the sample intervals. The digital files were imported into spreadsheet files that were verified back to the original assay files prior to being imported into the PDB.
14.4.1.5	QA/QC Data The Senior Geologist matched assays to sample numbers on the original log and extracted the QA/QC data from the sample stream into an excel spreadsheet which was used to monitor QA/QC data independently from the sample stream. The QA/QC data was sorted into worksheets for blanks, standards and duplicates. The entire file was cross-checked between the Project Manager’s copy and the Senior Geologist’s copy. Once satisfied that the data was free of error it was sent via email to Dr. Smee for final analysis.
14.4.2	Data Storage The PDB was established and maintained using Gemcom modeling software. The PDB was backed up on CD-R media on a regular basis for the duration of the field program. Once the PDB was fully updated with all the results collected during the 2002-2003 programs, the data was copied and forwarded to Kinross Technical Services in CD-R format. This database is referred to as the Final Database (FDB). In addition to the digital files, all available hard copy data was organized and inventoried on a per hole basis. This information was copied in triplicate with one copy forwarded to Kinross Technical Services, one copy forwarded to Bema and a final copy stored at the Refugio Mine. Original hard copy data was packaged and stored at the CMM office in Copiapo.
14.4.3	Data Management and Security Only the QA/QC Database Manager had the necessary file access permissions to update the PDB. Only Kinross’ Manager of Technical Services could access and modify the FDB.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

14-48

Compañia Minera Maricunga

15.0	ADJACENT PROPERTIES
	The Refugio Project is located within the Maricunga Gold Belt of northern Chile. Since 1980, a total of 40 million ounces of gold have been defined in the gold belt, (Muntean and Einaudi, 2000). The Maricunga Gold Belt hosts numerous mineral deposits of economic interest including:
	•	La Coipa, an epithermal gold-silver mine operated as a 50:50 joint venture between Placer Dome Inc and Kinross,
	•	Marte-Lobo, a gold porphyry deposit that was partially developed by Teck Corporation and,
	•	Cerro Casale a large (1.035 billion tonne), undeveloped, porphyry copper- gold deposit that is a joint venture between Placer Dome (51%), Bema (24%) and Arizona Star Resource Corporation (25%).

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

15-49

Compañia Minera Maricunga

16.0	MINERAL PROCESSING AND METALLURGICAL TESTING
	The 2002-2003 Exploration Program included the following metallurgical and physical property test work:
	•	cyanide soluble gold analyses
	•	cyanide soluble copper analyses
	•	direct agitated cyanidation tests (bottle roll)
	•	column leach tests
	•	Bond work index tests
	•	specific gravity determinations
	•	Acid Based Accounting (ABA) tests

16.1	Cyanide Soluble Assays
	Both gold and copper cyanide soluble analyses were routinely analyzed for both Verde and Pancho. Results were used to guide selection of samples for bottle roll and column leach tests. Cyanide copper analyses were also used to adjust process operating costs during pit optimization of the resource model. The mechanics of the Process Cost Adjustment Factor (PCAF) are discussed elsewhere in this report.
16.2	Direct Agitated Cyanidation (Bottle Roll) Tests
	Bottle roll composite samples were selected from the half core library based on lithology, oxidation state, grade and spatial location. A total of 43 bottle roll samples were selected from the Verde with an additional 15 samples collected from Pancho. The samples were shipped to McClelland Laboratories, Nevada for testing.
	The grade selection criteria used for selection of the bottle roll tests were:

•	Low grade	0.50  to  0.75 g/t
•	Mid-grade	0.75  to  1.0 o  g/t  and
•	High grade	> 1.0 g/t

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

16-50

Compañia Minera Maricunga

Oxidation state was determined based on the visual logging of oxidation noted in the geological logs. Three distinct oxidation levels were defined.

•	Oxide	> 90 %  oxidation of sulphide
•	Mixed	between 10% and 90% oxidation of sulphides and
•	Sulphide	< 10 %  oxidation of sulphides.

Samples were selected from the four major ore bearing lithologies, post mineral intrusives and Laguna tuffs were not sampled.

Sample selection was restricted within an optimized pit shell generated at a US $400 gold price. Table 16-1 summarizes the Bottle Roll sample selection.

Table 16-1 Summary of Bottle Roll Sample Selection

Deposit	Grade
	Low (0.50 - 0.75)	Medium  (0.75-1.00)	High (>1.00 g/t)
Pancho     # of samples  Verde W  # of samples  Verde E    # of samples	5 5 9	5  5  5	5  11  10
Deposit	Oxidation
	Oxide (> 90%)	Mixed  (10-90%)	Sulphide  (<10%)
Pancho     # of samples  Verde W  # of samples  Verde E    # of samples	6 8 9	3  5  6	6  8  9
Deposit	Lithology
	Diorite	Breccia	Mineralized Post Mineral Intrusive	Dacite Porphyry
Pancho     # of samples  Verde W  # of samples  Verde E    # of samples	6 - -	9  7  9	- 5 6	-  9  9

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

16-51

Compañia Minera Maricunga

Results of the bottle roll tests show distinct differences in recovery based on the level of oxidation. Oxide ores exhibit higher recoveries than Mixed ores which, in turn, demonstrate higher recoveries than the Sulphide ores. These finding agreed with the historical metallurgical testwork and experience gained from mining in the pit.

A summary of the bottle roll test results is presented in Table 16-2

Table 16-2 – Summary of Bottle Roll Results

Deposit	Oxidation Class	Recovery (%)
Verde	OXIDE  MIXED  SULPHIDE	72.3  63.8  53.0
Pancho	OXIDE  MIXED  SULPHIDE	83.0  72.8  68.6

16.3 Column Leach Tests Results of the bottle roll tests were used to target specific areas of the deposit with the HQ diameter drill holes that were used to collect the sample material necessary for column leach analysis. Composite samples were collected from whole HQ diameter core that was drilled specifically for the metallurgical test program. A total of 10 HQ diameter holes (1,500 meters) of core were drilled at Verde for the purpose of column leach tests. An additional seven (7) samples were collected at Pancho but unlike the Verde samples, the Pancho samples were taken from remaining half cores available at the time of sample selection.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

16-52

Compañia Minera Maricunga

For the Verde samples, column tests were established for three different product sizes (6.3, 9.5 and 15.8mm) . The same sample selection matrix established for the bottle roll tests guided column leach sample selection. Grade of the HQ diameter core was confirmed by cutting a sliver sample for each two-meter interval. Table 16-3 summarizes the Column Leach Sample program.

Table 16-3 Summary of Column Leach Sample Selection

Deposit	Grade
	Low (0.50 - 0.75 g/t)	Medium  (0.75 - 1.00 g/t)	High (>1.00 g/t)
Pancho     # of samples  Verde W  # of samples  Verde E    # of samples	2 2 2	2  2  2	3 1 2
	Oxidation
Deposit
	Oxide (>90%)	Mixed  (10-90%)	Sulphide (<10%)
Pancho     # of samples  Verde W  # of samples  Verde E    # of samples	3 - 3	1  -  3	3 5 -

Results of the column leach tests matched the bottle roll results. Table 16-4 summarizes the results of the column leach tests completed as part of this study.

Table 16-4 Summary of Column Leach Test Results

Deposit	Ore Type	Crush Size (k80mm)	Recovery  (%)
West Verde	Sulphide	15.8 9.5 6.3	44.0  53.7  57.2
East Verde	Mixed	15.8 9.5 6.3	70.5  73.6  79.3
East Verde	Oxide	9.5	77.0
Pancho	Sulphide	9.5	71.6
Pancho	Mixed	9.5	79.8
Pancho	Oxide	9.5	89.0

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

16-53

Compañia Minera Maricunga

The recovery results presented represent final recovery numbers after all the test work analysis was completed. At the time of the reserve estimate, final recovery estimates were not available. As a result, the recoveries used to generate optimum pit shells were slightly different than those noted above. A comparison of the recovery rates used for optimization purposes and those resulting from the final testwork are presented in Table 16-5.

Table 16-5 Comparison of Optimization versus Test Work Recovery

Deposit	Ore Type	Optimization  Recovery  (%)	Bottle Roll  Recovery  (%)	Column Test  Recovery  (%)
Verde	Oxide  Mixed  Sulphide	67                             67                             48	72                     64                     53	77                           74                           54
Pancho	Oxide  Mixed  Sulphide	80                             75                             70	83                     73                     69	89                           80                           72

On receipt of the final recovery results, the pit optimizations for Verde and Pancho were re-run to determine what impact the change in recovery had on the reserve estimates. Based on analyses of optimum pit shells, it was determined that the change in recovery had minimal impact on the pit designs and reserve estimates of the deposits.

Figures 16-1 and 16-2 are plan views of the Verde and Pancho pits showing the metallurgical sample distribution within the design pit shells.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

16-54

Compañia Minera Maricunga

Figure 16-1 Metallurgical Sample Locations – Verde Pit – 4300 Level

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

16-55

Compañia Minera Maricunga

Figure 16-2 Metallurgical Sample Locations – Pancho Pit – 4300 Level

16.4 Bond Work Index Bond Work Index (BWI) samples were collected from the 10, HQ diameter core holes drilled at Verde. Bond work index tests required selection of whole HQ core pieces of approximate length of 0.15 – 0.20 meters. For Verde, such pieces were selected every 4 meters downhole. All ten Verde HQ diameter holes were sampled in this manner, a total of 296 samples. At Pancho, BWI samples were collected from the standard diameter drill holes. Whole core pieces were taken approximately every 20 meters down hole. A total of 67 samples were selected. All samples were shipped to Hazen Research, Colorado, for analysis.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

16-56

Compañia Minera Maricunga

Results of the BWI tests indicated limited variability in the rock mass therefore, the BWI data was not modeled when completing the Verde and Pancho resource models

16.5

Specific Gravity As part of the BWI tests above, Hazen completed specific gravity analyses on all the BWI samples submitted. In addition to this data, 1171 samples of half core were selected from the half core library. Samples, roughly 15 cm in length, were selected approximately every 30 meters downhole throughout the entire campaign. The samples were shipped to ACME laboratories, Santiago where they were analysed using the wax dipped immersion method. A summary of the specific gravity results is presented in Table 16-6.

Table 16-6 – Specific Gravity Results

Deposit	Unit	SG
Verde  Verde  Verde  Verde  Verde  Verde	Oxide  Mixed  Sulphide  Laguna Tuff  Post Mineral Intrusive  Waste Dump Material	2.45  2.52  2.57  2.00  2.60  2.00
Pancho  Pancho  Pancho	Oxide  Mixed  Sulphide	2.27  2.38  2.49

Both Pancho and Verde exhibit trends of increasing Specific Gravity with depth, leading to the decision to model the deposit Specific Gravity by the oxidation surface rather than by rock type.

16.6

Acid Rock Drainage Study At the request of the CMM environmental department (Jorge Herrera), ALS Chemex La Serena sent 71 pulp samples to ALS environmental laboratories Santiago for acid based accounting testwork. A subset of 40 of these was proposed for testwork on June 6. The forty samples selected were from widely spaced locations in the anticipated pit designs for both Verdes and Pancho. Samples included both ore and waste in major rock types in all redox zones.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

16-57

Compañia Minera Maricunga

17.0	MINERAL RESOURCE AND MINERAL RESERVE ESTIMATION
17.1	General
	Mining experience at Verde demonstrated the existence of hard boundaries differentiating lithological units that are barren from those hosting mineralization. It also demonstrated the importance of an accurate recovery model. Modeling these features had been hampered in the past by the relatively limited amount of diamond drilling completed on the project. The 2003 resource block model incorporates lithological and oxidation models to better define grade distribution and gold recovery.
17.2	Lithological Model
	The resource models rely on the lithological interpretation completed by the site geologists to discriminate between mineralized and barren lithologic units. The lithology model, developed from cross sectional, longitudinal and level plan interpretation, identified distinct lithological boundaries that were used as hard boundaries during grade interpolation. The lithology model was developed largely from the observations taken from the 2002 –2003 drill core and is restricted to the area below the current topographic surface. The lithological model was not extended above the topographic surface.
17.3	Recovery Model
	Metallurgical recovery is a key aspect in modeling the Verde and Pancho deposits. Significant effort was directed towards establishing a relationship between metallurgical recovery and the level of oxidation visually estimated by the site geologists. As with construction of the lithology model, the recovery model was developed by interpretation of cross section, longitudinal section and level plan data by the site geologists. The interpreted surfaces were used as hard boundaries during pit optimization. The surfaces are highly irregular, producing an undulating surface similar to what has been observed in the pit walls at Verde. Figure 17-1 presents a typical section through the Verde East deposit showing the interpreted oxidation surface and drill grades.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-58

Compañia Minera Maricunga

Figure 17-1 Verde East – Section 5NW

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-59

Compañia Minera Maricunga

17.4	Grade Model
17.4.1	Sample Compositing Sample compositing was not used in completing the 2003 resource models. All grade interpolation was based on the raw assay intervals, which were generally 2.0 meters in length.
17.4.2	Statistical and Geostatistical Analyses
17.4.2.1	Excluded Data The following data was not used in completing the 2003 resource block models:
	•	blast hole data,
	•	trench data and,
	•	drill data within the previously mined portion of the Verde pits.
	Unexplained missing data was assigned a zero value for all analyses. Explained missing data was treated as missing data and ignored during processing.
17.4.2.2	Grade capping Grade capping was examined for gold, copper and cyanide soluble copper data and was determined based on analysis of the statistics of individual lithological units and oxidation classes. Capping of gold values affects 0.2% of the total sample population resulting in a decrease in modeled gold content of less than 1%. Capping factors for gold ranged from 6.0 to 6.5 g/t at Verde to 8.0 to 9.0 g/t at Pancho.
17.4.2.3	Variography Directional semi-variograms, (15o increments horizontal, 30oincrements vertical) were modeled for both Verde and Pancho, examining only the drill data below the current topographic surface. The variography was examined independently for each lithologic and oxidation zone modeled. The models generally indicated anisotropic, two-structure models with well defined nugget values that were less than half the modeled sill value. The models show good ranges in all directions, typically in excess of 100 meters.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-60

Compañia Minera Maricunga

17.4.3	Model framework
	Both the Verde and Pancho block models were rotated parallel to the direction of drilling. At Verde, the block model is rotated 60 degrees NE about the Z-axis. At Pancho, the angle of rotation is 38°NE around the Z-axis. Both models have sufficient extent to encompass all known mineralization. The model cell size at Verde was set at 20 x 20 x 10 meters (X, Y and Z direction) while at Pancho, the model cell size was decreased to 10 x 10 x 10 meters.
17.4.4	Gold Grade Interpolation
	The primary interpolation method for gold was ordinary kriging which used the anisotropy and ranges supported by the best variogram model for each lithologic and oxidation zone. Grade interpolation was verified by inverse distance to the 2ndpower and nearest neighbour interpolation. Both capped and uncapped grade estimates were completed. Grade interpolation utilized octant search restrictions, requiring a minimum of two octants to estimate the grade of a block. The minimum number of samples per octant was 1 with a maximum of 4. A minimum of 4 samples and maximum of 24 samples were required to estimate the grade of any block. The number of samples from one drill hole that could influence a block grade estimate was limited to six (6).
17.4.5	Copper Interpolation
	Copper grades were not modeled at Verde, as results of the metallurgical testwork did not indicate that copper was impacting grade recovery or operating costs. At Pancho, analysis of the drill data indicated that copper grade was closely related to the oxidation class. Therefore, copper grades were interpolated into the Pancho model blocks independently of the gold grade. Two interpolation methods were used to estimate copper grade, an inverse distance to the 2ndpower interpolation method was used as the primary estimation method. This was verified using an ordinary kriging interpolation method. Search ellipsoid anisotropy and ranges were based on the copper variograms for each recovery class.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-61

Compañia Minera Maricunga

17.4.6	Cyanide Soluble Copper Interpolation
	Cyanide soluble copper (CN Cu) was not modeled at Verde. At Pancho, analysis of the drill data indicated that CN Cu was closely related to the oxidation state. CN Cu grade interpolation was completed for each the three oxidation zones separately, using only cyanide soluble copper data extracted (coded) for each oxidation class to estimate grade within the class. Two interpolation methods were used, an inverse distance to the 2nd power interpolation was the primary interpolation method with a nearest neighbour interpolation method completed to verify the primary grade interpolation method. Search ellipsoid anisotropy and ranges were determined based on the results of the variograms of each recovery class.
17.4.7	Model Checking and Verification
	Block model verification included visual inspection of the model after grade interpolation was completed to examine if the interpolation honoured hard boundaries and grades were consistent with the available raw data. The model was manually reviewed in plan and section using Datamine visualization tools. Visual inspection did not identify any errors in the model. A second check involved comparison of the various grade interpolation results. Comparison of the grade tonnage curves for the various estimation methodologies does not indicate significant differences in tonnage, or average grade. Table 17-1 provides comparisons of the Verde model grades using the three different grade interpolation parameters. Similar results were produced at Pancho.

Table 17-1 Comparison of Verde Interpolation Methods

Measured and Indicated
COG	Kriged Estimate			Inverse Distance			Nearest Neighbour
	tonnes (x 1000)	grade  (Au g/t)	Contained  Au (ozs)	tonnes (x 1000)0	grade  (Au g/t)	Contained  Au (ozs)	tonnes (x 1000)	grade  (Au g/t)	Contained   Au (ozs)
0.10  0.20  0.30  0.40  0.50  0.60  0.70  0.80  0.90	328,190 292,525 256,331 217,153 177,278 141,124 105,693 77,251 55,640	0.60  0.65  0.71  0.78  0.85  0.93  1.02  1.12  1.22	6,331,000   6,113,000   5,851,000   5,446,000   4,845,000   4,220,000   3,466,000   2,782,000   2,182,000	327,233 291,096 254,344 215,556 176,584 139,988 105,567 76,787 56,124	0.60  0.66  0.72  0.78  0.85  0.93  1.03  1.13  1.24	6,312,000   6,177,000   5,888,000   5,406,000   4,826,000   4,186,000   3,496,000   2,790,000   2,237,000	306,462 264,421 221,882 182,567 147,079 117,411 93,538 74,846 61,008	0.63  0.71  0.80  0.90  1.01  1.13  1.25  1.38  1.50	6,207,000  6,036,000  5,707,000  5,283,000  4,776,000  4,266,000  3,759,000  3,321,000  2,942,000

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-62

Compañia Minera Maricunga

Comparison of the resource model to historical estimates indicates that the 2003 Verde model contains more tonnes at almost exactly the same average grade as what was predicted by MRDI’s 1998 model. Table 17-2 presents a comparison of the MRDI 1998 Verde model to the 2003 Verde estimate.

Table 17-2 Comparison of 2003 Estimate to 1998 MRDI Estimate

	2003 Estimate			Net Change to 1998 MRDI Model			MRDI 1998 Estimate
Cut off Grade  (Au g/t)	Tonnes (t x 1,000)	Grade (Au g/t)	Gold (ozs)	Tonnes (%)	Grade (%)	Gold (%)	Tonnes (t x 1,000)	Grade (Au g/t)	Gold    (ozs)
Measured and Indicated
0.2  0.4	292,625  218,550	0.66  0.78	6,191,000  5,467,000	33% 56%	13% 5%	51% 64%	219,300  140,520	0.58  0.74	4,089,000  3,343,000
0.5	179,340	0.85	4,901,000	63%	4%	69%	110,080	0.82	2,902,000
0.7  1.0  1.5	107,449  40,975  8,865	1.02  1.33  1.85	3,517,000  1,747,000  526,000	68% 74% 179%	3% 5% 9%	73% 84% 202%	63,890  23,490  3,180	0.99  1.26  1.70	2,034,000  952,000  174,000
Inferred
0.2  0.4	114,533  73,248	0.52  0.64	1,915,000  1,510,000	121% 165%	9% -6%	141% 150%	51,880  27,620	0.48  0.68	793,000  603,000
0.5	53,361	0.71	1,222,000	181%	-5%	166%	19,020	0.75	460,000
0.7  1.0  1.5	21,101  4,347  460	0.90  1.20  1.79	610,000  168,000  26,000	130% 40% 35%	-1% -4% -1%	127% 33% 30%	9,190  3,100  340	0.91  1.26  1.80	269,000  126,000  20,000

Table 17-3 presents the historical estimates at Pancho as compared to the 2003 estimate.

Table 17-3 Historical Pancho Resource Estimate Comparison

Source	Date	Classification	Cut off grade  (Au g/t)	Tonnes  (t x 1,000)	Grade  (Au g/t)
Bema  Kinross  Bema	pre-1997  1998  1998	unclassified  unclassified  unclassified	0.62  0.60   0.60	92,737  122,733  109,380	0.86  0.91  0.85
Kinross	2003	measured, indicated and inferred	0.60	160,082	0.82

As at Verde, the 2003 Pancho estimate indicates a tonnage increase while the average grade remains very close to the grade predicted by the historic estimates..

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-63

Compañia Minera Maricunga

17.5 Resource Classification The mineral resource estimates for Verde and Pancho have been prepared and classified in accordance with the Canadian Institute of Mining, Metallurgy and Petroleum’s (“CIM)Standards on Mineral Resources and Reserves, Definition and Guidelines.Resource classification was based on the variogram ranges for each litho-recovery zone. Table 17- 4 presents the classification matrix used in estimating the 2003 resource models for Verde and Pancho.

Table 17-4 Resource Classification Matrix

Zone	Domain	Axis	Measured  Searcn Radii (m)	Indicated Searcn Radii (m)	Inferred Searcn Radii (m)
Verde	31& 32	X Y Z	30  25  25	60  50  50	120  100  100
	21& 22	X Y Z	30  30  25	60  60  50	120  120  100
	11& 12	X Y Z	25  25  30	50  50  60	100  100  120
	13,23 & 33	X Y Z	30  30  25	60  60  50	120  120  100
Pancho	1	X Y Z	30  30  10	60  60  20	120  120  40
	2	X Y Z	12.5  12.5  10	25  25  20	50  50  40
	3	X Y Z	30  45  10	60  90  20	120  180  40
	4	X Y Z	25  40  10	50  80  20	100  160  40
	5	X Y Z	22.5  22.5  10	45  45  20	90  90  40

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-64

Compañia Minera Maricunga

17.5.1	Measured Resource
	Blocks where the gold grades are interpolated from at least 4 samples located in two different octants and sourced from a minimum of two drill holes within the first search ellipsoid limits (Measured) defined in Table 17-4.
17.5.2	Indicated Resource
	Blocks where the gold grades are interpolated from at least 4 samples located in two different octants and found within the second search ellipsoid limits (Indicated) defined in Table 17-4.
17.5.3	Inferred Resource
	Blocks where the gold grades are interpolated from at least 4 samples sourced from a minimum of one drill hole located in two different octants and found within the third search ellipsoid limits (Inferred) defined in Table 17-4.
17.6	Resource Estimate
17.7	General
	It is Kinross Policy that resource estimates for open pit mine projects are reported within optimized or designed pit shells at assumed gold prices that represent reasonable, long-term price projections. Final pit designs have been completed for both Verde and Pancho and as a result, only Proven and Probable reserves are reported.
	As per CIM Guidelines, pit optimization only considers resource blocks classified as Measured and Indicated when evaluating project economics. Accordingly, only Measured and Indicated resource blocks were selected for pit optimization. Inferred blocks were excluded from the optimization algorithm by manually assigning a gold content of zero to all Inferred blocks.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-65

Compañia Minera Maricunga

17.8 Pit Optimization All pit optimization was completed using Whittle 4X which utilizes the Lerchs-Grossman optimization algorithm. Pit optimization of the Verde and Pancho models assumed metallurgical recoveries as noted in Table 16-5. Operating costs were developed from first principles and verified to historic operating costs. Table 17-5 summarizes the pit optimization assumptions used to develop the optimized pit shells.

Table 17-5 – Summary of Pit Optimization Parameters

Verde	Mining cost ($/t)  Process cost ($/t)  Mining recovery (%)  Dilution (%)  Slope Angle (degrees)	$0.91 $3.14 95% 5% 45
Pancho	Mining cost ($/t)  Process cost ($/t)  Mining recovery (%)  Dilution (%)  Slope Angle (degrees)	$1.21 $3.27 95% 5% 45

Optimization of the Pancho resource model also included a Process Cost Adjustment Factor (PCAF) to reflect higher process operating costs resulting from increased cyanide consumption due to higher CN Cu grades. The PCAF equation used for optimization of the Pancho resource is:

PCAF = ((CN Cu x $0.0035) + Process Cost))/Process Cost

The equation is based on the assumption that 3.5 grams of cyanide were required for every 1 ppm of CN Cu. Cyanide costs were estimated at $1.00 per tonne.

The pit optimization process produced a series of nested pit shells at various gold prices by applying a revenue factor that varies the input gold price. For the Verde and Pancho optimizations, the revenue factor ranged from a low of 0.2 ($70 per oz.) to a high of 2.0 ($700 per oz) in increments of 0.05, producing 37 nested pit shells for each deposit, each of which was evaluated based on the net revenue of each shell.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-66

Compañia Minera Maricunga

	Typically, the shell with the highest net revenue is selected to guide the final pit design.
17.9	Final Pit Design Final pit design for Verde and Pancho assumed the following:
	•	a 10 meter bench height;
	•	bench face angles of 65°to 70°;
	•	berm widths of 8 to 11 meters;
	•	berm interval of 20 meters;
	•	inter-ramp angles of 38°to 52.5°and
	•	haul road gradient at 10% with a 25 meter road width.
	Final pit designs were developed using Datamine. The final pit designs were used to generate triangulated surfaces representing the end of mining topography. These surfaces were used to extract the reserve tonnages and grades on a bench-by-bench basis from the respective block models. The extracted block model summaries were then diluted and mining losses were applied. In estimating the reserves, it was assumed that the 5% dilution would negate the 5% loss. As a result, ore tonnage would not change but the average grade would be reduced. Dilution was applied on a bench-by-bench basis. Dilution grade was estimated from the average grades of all samples below the ore cut-off grade as determined by the pit optimization.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-67

Compañia Minera Maricunga

17.10 Mineral Resource and Reserve Estimate Table 17-6 presents the Proven and Probable Reserves for the Verde and Pancho deposits as of December 1, 2003 at a gold price of $350 per ounce which represents the Base Case gold price for resumption of mining activity.

Table 17-6 Proven and Probable Reserves @ $350 per ounce

Deposit	Class	Tonnes (t x 1,000)	Grade (Au g/t)	Gold (ounces)
Verde East	Proven  Probable  Proven and Probable	23,235  6,522  29,757	0.86  0.78  0.84	642,000  164,000  806,000
Verde West	Proven  Probable  Proven and Probable	36,569  11,252  47,821	0.93  0.82  0.90	1,093,000  304,000  1,397,000
Subtotal	Proven  Probable  Proven and Probable	59,804  17,774  77,578	0.90  0.82  0.88	1,735,000  468,000  2,203,000
Pancho	Proven  Probable  Proven and Probable	38,516  7,960  46,476	0.84  0.75  0.82	1,040,000  190,000  1,230,000
Total	Proven  Probable  Proven and Probable	98,320  25,734  124,054	0.88  0.80  0.86	2,775,000  658,000  3,433,000

Table 17-7 presents the Proven and Probable Reserves for the Verde and Pancho deposits as of December 1, 2003 at a gold price of $325 per ounce.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-68

Compañia Minera Maricunga

Table 17-7 Proven and Probable Reserves @ $325 per ounce

Deposit	Class	Tonnes (t x 1,000)	Grade (Au g/t)	Gold (ounces)
Verde East	Proven  Probable  Proven and Probable	21,583  5,351  26,934	0.85  0.76  0.83	587,000  131,000  718,000
Verde West	Proven  Probable  Proven and Probable	34,850  9,671  44,521	0.91  0.79  0.88	1,015,000  245,000  1,260,000
Subtotal	Proven  Probable  Proven and Probable	56,433  15,022  71,455	0.88  0.78  0.86	1,602,000  376,000  1,978,000
Pancho	Proven  Probable  Proven and Probable	23,061  4,615  27,676	0.90  0.81  0.89	669,000  123,000  792,000
Total	Proven  Probable  Proven and Probable	79,494  19,637  99,131	0.89  0.79  0.87	2,271,000  499,000  2,770,000

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

17-69

Compañia Minera Maricunga

18.0

INTERPRETATION AND CONCLUSIONS

The Mineral Resource and Reserve Estimates for the Refugio Project have been prepared under the direct supervision of Rodney Cooper, P.Eng., Kinross Gold Corporation’s Director of Technical Services and Wes Hanson, P.Geo., Kinross Gold Corporation’s Manager of Technical Services. Both individuals are “qualified persons” as defined by Canada’s National Instrument 43-101. Messrs. Cooper and Hanson are satisfied the data used in the estimation of Mineral Resources and Reserves is free of gross errors and omissions and is of suitable quality and sufficient quantity to use in estimating resources and reserves for the Verde and Pancho deposits.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

18-70

Compañia Minera Maricunga

19.0

REFERENCES:

Anonymous, 1991, Final feasibility study, Verde Gold Deposit Refugio Property, Northern Chile: Mineral Resources Development Inc., unpublished report for Bema Gold Corp. and Compania Minera Refugio, several volumes. Anonymous, 1992, Revised feasibility study, Verde Gold Deposit Refugio Property, Northern Chile: Mineral Resources Development Inc., unpublished report for Compania Minera Maricunga, Bema Gold Corp. and Amax Gold Inc., one volume. Ardilla, R.J., 1997. Summary Of The Geology of The Pancho Gold Deposit, Refugio. Internal Report, Minera Cyprus Chile Ltda. CMM Staff Geologists, Carrasco, M, Crawford, K, Duran, M. Rodriquez, B. Romero, I, Strusvieci, M, Urrutia, H. Cortes, 2001, Geologia del Yacimento Verde (Geology of the Deposit Verde,English Translation): Compania Minera Maricunga, unpublished geologic map (1:2,000). NOTE: The mapping was conducted in 2000 and 2001 after the mining was completed; therefore, it is a“final pit map”.The map is of good quality and the mapping is generally confirmed by the Kinross mapping and compilation conducted in 2003 as part of the feasibility study. AutoCad File: geologia2001.dwg. Compania Minera Maricunga, Procedimiento Mapeo Pozos Tronaduras. Procedures Files, in CMM files. Cortes, C., Omar 2000, Informe de Pruebas Geolmetalurgicas de Mineralizacion Mixta. Noviembre 1999 - Febrero 2000, In CMM files. Emmons, D, Verde East and West Geologic Mapping, 2003, Kinross internal report, May 10,2003.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

19-71

Compañia Minera Maricunga

Flores V., Roman, undated, Precious Metal Deposits of the Refugio Area, Northern Chile, In Kinross Files.

Khin, B.G. 2003. Petrographic Descriptions, Pancho Project, Chile. Consultants Report Prepared For Kinross Gold Corp.

Kinross Technical Services, 1998, Refugio Property Review, June 1998, report in Kinross Files.

Magri, Eduardo and NCL Consulting, Drillhole Database, Audit, SD001 a SD177., , July 1995. in files Refugio Mine Site

McDermott, Ryan, 2000, Refugio Ore Type Classification, Memorandum to Brian Scott October 12 2000, in Kinross files.

McDermott, Ryan, 2001, Refugio Model Summary and Explanation, Memorandum to Rick Dye, January 30 2001, in Kinross files.

Miller, Vic, and McDermott, Ryan, 1999, Refugio Trip, Memorandum to Rick Dye Dated September 14, 1999, in Kinross files.

Meister, Susan N., 1998, Verde Block Model and Resource, Estimate Report, dated December 16, 1998. In CMM files.

Miller, Victor, 2002, Refugio Resources and Reserves, Memorandum to Rick Dye dated February 20 2002, in Kinross files.

Muntean, J. L., 1998, Magmatic-hydrothermal gold deposits of the Maricunga Belt, Northern Chile: Stanford University, Ph. D. dissertation, 407 p.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

19-72

Compañia Minera Maricunga

Muntean, J. L. and Einaudi, M. T., 2000, Porphyry gold deposits of the Refugio District, Maricunga Belt, Northern Chile: Economic Geology, v. 95, pp. 1445-1472.

Muntean, J.L. and Einaudi, M.T. 2001. Porphyry Epithermal Transition: Maricunga Belt, Northern Chile. Economic Geology, v 96, pp 743-772.

Pritting, J, Geologic Summary of the Pancho Project, Refugio District, Chile, May 2003.

Romero, Ivan, 2000, Informe de Cierre Temporal de Actividades Cia Minera Maricunga, Noviembre de 2000, Modificacado enero 9 2001. In CMM files

Romero, Ivan, 2001, Programa de Termino de Geologia y Planificacion Mina. In CMM files

Scott, Brian, 1998, Trip to Refugio-Jan 12-22, 1998, Bema Gold Corporation Memorandum to Tom Garagan, in CMM files.

Scott, Brian, 1999, Deep Verde Drilling, Consolidated Westview Resources Memorandum to Tom Garagan, dated April 23, 1999, in CMM files.

Scott, Brian, 1999b, Refugio Metallurgical Info 1991 - 1998, Bema Gold Corporation Memorandum to Tom Garagan, dated August 31. in CMM files.

Scott, Brian, 2000, Refugio - Met Program 2000 - Geologic Observations, Bema Gold Corp. Memorandum to Tom Garagan, in CMM files.

Sillitoe, R.S., 1991. Gold Metallogeny of Chile, An Introduction. Economic Geology v.86, pp 1187-1205.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

19-73

Compañia Minera Maricunga

Sillitoe, R. H. and Camus, F., 1991, A special issues devoted to gold deposits in the Chilean Andes: Economic Geology, v. 86, no. 6. NOTE: contains papers on the Refugio gold deposits and the Maricunga Gold Belt.

Spiteri, J. G. and Cohoon, G. A., 2002, Report on resources Verde deposits Refugio, Chile (draft): Spiteri Geological and Mining Consultants Inc., unpublished report for Kinross Gold Corp., 101 p.

SNC-Lavalin, Pre-Feasibility Study, Refugio Project, September, 2003

SNC-Lavalin, Refugio Mine Metallurgical Summary, December, 2003

Strusievici, Mihai 1998, A Review of the Geology of Verde Gold Deposit, Preliminary Conclusions - Implication in the Resource Model, in CMM files

Strusievici, M., 1999, Geological report, mineralization model of Verde – targets for mine site exploration: Compania Minera Maricunga, unpublished report with geologic map, 24 p. NOTE: Describes the staff geologic mapping for the 1999 Verde pit maps.

Strusievici, M., et al., 1999, Geologic maps Verde deposit: Compania Minera Maricunga, unpublished geologic map (1:1,000). NOTE: The map is the product of a 1998 pit mapping project by CMM staff including geologists Crawford, K., Duran, Mauricio, and Urrutia, Hegel. The map was compiled and interpreted by Strusievici and others. AutoCAD File: Verde Geo.dwg

Vila, T. and Sillitoe, 1991, Gold-rich porphyry systems in the Maricunga Belt, Northern Chile: Economic Geology, v. 86, pp. 1238-1260. NOTE: The paper has a good summary of the exploration history of the Maricunga Belt.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

19-74

Compañia Minera Maricunga

Williams, S.A., 1997. Petrographic Descriptions, Pancho Project, Refugio District. Consultant’s Report Prepared For Minera Cyprus Chile, Ltda.

Wolverson, N.J. 1997. Pancho Summary Report, 1997. Internal Report, Compania Minera Maricunga.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

19-75

Compañia Minera Maricunga

20.0	ADDITIONAL REQUIREMENTS FOR TECHNICAL REPORTS ON DEVELOPMENT PROPERTIES
	The Refugio Mine is a former producing mine and significant infrastructure is already in place. The reserve estimates presented in this document are based on the assumption that CMM plans a capital expenditure of $101.1 million to upgrade the existing operation. The capital expenditure will repair, refurbish and replace existing equipment with the objective of increasing the daily production throughput to 40,000 tonnes per day for Verde ores, decreasing to 35,000 tonnes per day when the plant begins processing Pancho ores. Based on the 2003 Reserve Estimate, the current reserves are sufficient to support the operation for 9.6 years.
20.1	Mobile Equipment The proposed capital cost estimates includes $30.1 million to purchase a new mining fleet. Fleet requirements were estimated from first-principles based on the proposed mine plan. A partial list of the major mine equipment necessary to support the planned production rate includes:
		3 – CAT 994D Front End Loaders 8 – CAT 785C Haul Trucks 2 – Ingersol Rand 250 mm Rotary Drills 2 – CAT D9R Bulldozers 1 – CAT D8R Bulldozer 2 – CAT 16H Motor Graders 1 CAT 834 Wheel Dozer 1 50t Water Truck 1 CAT 330CL Backhoe 1 Ingersol Rand 100mm Boom Drill

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

20-76

Compañia Minera Maricunga

		The capital cost estimates includes costs for support equipment including fuel and lube trucks, flatbed trucks, portable light towers, and service and welding trucks.
20.2	Plant Infrasturcture A capital expenditure of $57.4 million is planned to upgrade the existing plant facilities. Key aspects of the upgrade include establishing a 110 km power line to replace the generated power currently at site. Additional capital expenditure is planned to upgrade the current crushing and conveying system, replacing certain components to improve availability and throughput. Capital costs of the upgrades were estimated from first principles based on review of the current design drawings and costs estimates provided by a major North American Engineering Consultant. Another North American Engineering Consultant reviewed the capital cost estimates and found them to be accurate.
20.3	Life-of-Mine Plan A complete life-of mine-plan was prepared based on the estimated reserves. The Life-of-Mine Plans is presented in Table 20-1 below.

Table 20-1 Refugio Project Life-of-Mine Plan.

	Year 1	Year 2	Year 3	Year 4	Year 5
Tonnes processed (kt)	12,000	14,000	14,000	14,000	14,000
Head grade (Au g/t)	0.90	0.87	0.78	0.85	0.88
Recovery (%)	67.1%	64.3%	66.5%	66.5%	60.4%
	Year 6	Year 7	Year 8	Year 9	Year 10
Tonnes processed (kt)	14,000	10,900	10,500	10,500	10,154
Head grade (Au g/t)	0.87	0.80	0.89	0.86	0.89
Recovery (%)	59.1%	76.7%	76.1%	74.1%	73.8%

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

20-77

Compañia Minera Maricunga

	The Life-of-Mine plan was used to establish production requirements for manpower and equipment. Mine operating costs were developed from first principles based on the mining rates indicated in the Life-of-Mine plan.
20.4	Environmental Considerations
	No significant permitting issues have been identified. Most of the proposed plant improvements represent little to no change to the existing footprint. The increased reserves will result in the need to permit additional leach pad capacity but this is not considered to be a risk, especially as the existing permitted space is sufficient for the majority of the remaining reserves.
	Aside from leach pad capacity, the only other significant change affecting permitting is the proposed power line. Again, preliminary discussion with site environmental staff indicate that obtaining the necessary permits within a reasonable time frame is not considered to be a significant risk.
20.5	Capital and Operating Costs
	Capital and operating cost estimates were prepared based on first-principles and verified to historical mine operating costs available for the project. Operating costs are consistent with past operating experience and are considered to be an accurate representation of the costs anticipated once the proposed improvements have been completed to upgrade the existing infrastructure.
	A major North American Engineering firm estimated all capital costs of the project. A second independent engineering verified the capital cost estimate was reasonable. No significant errors or omissions were identified.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

20-78

Compañia Minera Maricunga

20.6 Financial Analysis A financial analysis has been completed based on the Life-of-Mine plan and the capital and operating cost estimates. The model is not a public document and is not included in this report. Review of the financial analysis indicates positive cash flow at gold prices above $325 per ounce.

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

20-79

Compañia Minera Maricunga

21.0 AUTHORS CONSENT FORMS

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

21-80

Compañia Minera Maricunga

Certificate and Consent

To Accompany the Kinross Gold Corporation 2003 Refugio Technical Report

I, Rodney A. Cooper, P.Eng., residing at 1460 Golden Meadow Trail, Oakville, Ontario, Canada, L6H 3J3 do hereby certify that:

1.	I am a mining engineer employed by Kinross Gold Corporation, 40 King Street West, Toronto, Ontario, Canada, M5H 3Y2, in the capacity of Director, Technical Services.
2.	I am a graduate of Queen’s University (1980) with an honours B.Sc. Applied Science (Mining) degree and have practiced my profession continuously since 1980, with the exception of a two year period in 1982 to 1984.
3.	I am a graduate of the University of Toronto (1984) with a master’s degree in Business Administration (M.B.A.).
4.	I am a member in good standing of the Professional Engineers of Ontario, licence number 9302506.
5.	I am a qualified person as defined by Canadian National Instrument 43-101.
6.	I supervised the overall preparation of the Refugio Technical Report, including resource modeling, reserve estimation, mine planning and scheduling, mining equipment selection, mine capital and operating cost estimation, metallurgical test- work, geotechnical and hydrological engineering related to the three open pit mines at Refugio. I certify that, to the best of my professional judgment, qualified persons, as defined under Canadian National Instrument 43-101, performed the tasks completed under my supervision.
7.	I am not aware of any material fact or material change with respect to the subject matter of the technical report, which is not reflected in the technical report, the omission to disclose which makes the technical report misleading.
8.	I have read National Instrument 43-101 and Form 43-101F1, and the technical report has been prepared in accordance with this instrument.
9.	I hereby consent to the use of this report for submission to any Provincial regulatory authority in Canada.

Toronto, Canada	Rodney A. Cooper, P.Eng.,
December 22, 2003	Director, Technical Services,
	Kinross Gold Corporation

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

21-81

Compañia Minera Maricunga

Certificate and Consent

To Accompany the Kinross Gold Corporation 2003 Refugio Technical Report

I, Wesley C. Hanson, P.Geo., residing at 391 Devonshire Terrace, Ancaster, ON, L9G 4R4 do hereby certify that:

1.	I am a geologist employed by Kinross Gold Corporation, 40 King Street West, Toronto, Ontario, Canada, M5H 3Y2, in the capacity of Manager, Technical Services.
2.	I am a graduate of Mount Allison University (1982) with a B.Sc. Applied Science (Geology) degree and have practiced my profession continuously since 1982.
3.	I am a member in good standing of the Association of Professional Geoscientists of Ontario, licence number 1013.
4.	I am a qualified person as defined by Canadian National Instrument 43-101.
5.	I supervised the overall preparation of the Refugio Technical Report, including resource modeling, reserve estimation, mine planning and scheduling, mining equipment selection, mine capital and operating cost estimation, metallurgical test- work, geotechnical and hydrological engineering related to the three open pit mines at Refugio. I certify that, to the best of my professional judgment, qualified persons, as defined under Canadian National Instrument 43-101, performed the tasks completed under my supervision.
6.	I am not aware of any material fact or material change with respect to the subject matter of the technical report, which is not reflected in the technical report, the omission to disclose which makes the technical report misleading.
7.	I have read National Instrument 43-101 and Form 43-101F1, and the technical report has been prepared in accordance with this instrument.
8.	I hereby consent to the use of this report for submission to any Provincial regulatory authority in Canada.

Toronto, Canada	Wesley C. Hanson, P.Geo.,
December 22, 2003	Manager, Technical Services,
	Kinross Gold Corporation

Technical Report on the Refugio Gold Project – Maricunga District - Chile 

21-82