IMPORTANT NOTICE

This report was prepared as a National Instrument 43-101 Technical Report, in accordance with Form 43-101F1, for Nevsun Resources (Eritrea) Ltd., a subsidiary of Nevsun Resources Ltd. (Nevsun) by AMEC (Perú) S.A. a division of AMEC Americas Limited (AMEC). The quality of information, conclusions, and estimates contained herein is consistent with the level of effort involved in AMEC’s services, based on: i) information available at the time of preparation, ii) data supplied by outside sources, and iii) the assumptions, conditions, and qualifications set forth in this report. This report is intended to be used by Nevsun, subject to the terms and conditions of its contract with AMEC. That contract permits Nevsun to file this report as a Technical Report with Canadian Securities Regulatory Authorities pursuant to provincial securities legislation. Except for the purposes legislated under provincial securities laws, any other use of this report by any third party is at that party’s sole risk.

CONTENTS
1.0	SUMMARY			1-1
	1.1	Introduction		1-1
	1.2	Geology and Mineralization		1-2
	1.3	Data, QAQC, EDA and Verification Sampling		1-3
	1.4	Metallurgical Testwork		1-3
	1.5	Resource Estimation		1-4
	1.6	Recommendations		1-7
		1.6.1	Phase I Work Program	1-9
		1.6.2	Phase II Work Program	1-10
2.0	INTRODUCTION AND TERMS OF REFERENCE			2-1
	2.1	Introduction		2-1
	2.2	Terms of Reference		2-2
3.0	DISCLAIMER			3-1
4.0	PROPERTY DESCRIPTION AND LOCATION			4-1
	4.1	Location		4-1
	4.2	Land Tenure		4-1
	4.3	An Overview of Eritrea		4-5
		4.3.1	Introduction	4-5
		4.3.2	Geography and Infrastructure	4-5
		4.3.3	Modern History	4-6
		4.3.4	Demography and Government	4-7
		4.3.5	Mining Industry and Legislation	4-8
		4.3.6	Mineral Property Title	4-8
		4.3.7	Environmental Regulations	4-10
5.0	ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY			5-1
	5.1	Accessibility		5-1
	5.2	Climate		5-3
	5.3	Local Resources and Infrastructure		5-3
	5.4	Physiography, Flora and Fauna		5-4
6.0	HISTORY			6-1
7.0	GEOLOGICAL SETTING			7-1
	7.1	Regional Geology		7-1
	7.2	Structural Interpretation of Western Eritrea		7-5
	7.3	Mineral Deposits of Eritrea		7-5
	7.4	Property Geology		7-10

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	TOC i
	October 2004

		7.4.1	Stratigraphy	7-11
		7.4.2	Intrusives	7-20
		7.4.3	Structure	7-22
		7.4.4	Metamorphism and Weathering	7-26
8.0	DEPOSIT TYPES			8-1
	8.1	Noranda/Kuroko VMS Deposit Model		8-1
	8.2	Bimodal Siliciclastic VMS Deposit Model		8-3
9.0	MINERALIZATION			9-1
	9.1	Introduction		9-1
	9.2	Host Rock		9-2
	9.3	Deposit Dimensions and Morphology		9-2
	9.4	Oxide Zone		9-4
	9.5	Acid Zone		9-5
	9.6	Supergene Zone		9-6
	9.7	Primary and Primary Zn Zones		9-6
	9.8	Footwall Alteration		9-7
10.0	EXPLORATION			10-1
	10.1	Introduction		10-1
	10.2	Coordinates and Datums		10-3
	10.3	Topography and Grid Survey Control		10-3
	10.4	Geological Mapping and Related Studies		10-4
	10.5	Remote Sensing and Satellite Imagery		10-6
	10.6	Geochemistry		10-6
		10.6.1	Stream Sediment Sampling	10-6
		10.6.2	Soil Geochemical Sampling	10-8
	10.7	Trenching		10-11
	10.8	Ground Geophysics		10-11
		10.8.1	Electromagnetic (EM)	10-11
		10.8.2	Magnetometer	10-11
		10.8.3	Induced Polarization (IP)	10-13
		10.8.4	Gravity	10-13
	10.9	Airborne Geophysics		10-13
	10.10	Mineralogical and Petrographic Studies		10-15
	10.11	Bulk Density Determination		10-15
	10.12	Preliminary Metallurgical Studies		10-15
	10.13	Drilling		10-15
	10.14	Other Studies		10-16
11.0	DRILLING			11-1

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	TOC ii
	October 2004

	11.1	Introduction		11-1
	11.2	Diamond Drilling		11-1
		11.2.1	Collar Surveys	11-3
		11.2.2	Downhole Surveys	11-4
		11.2.3	Logging	11-7
		11.2.4	Photography	11-8
		11.2.5	Recoveries	11-9
		11.2.6	Geotechnical Logging	11-10
		11.2.7	Dry Bulk Density Measurement	11-11
		11.2.8	Results	11-13
	11.3	RC Drilling		11-14
		11.3.1	Collar Surveys	11-14
		11.3.2	Downhole Surveys	11-14
		11.3.3	Logging	11-15
		11.3.4	Recoveries	11-15
		11.3.5	Results	11-16
	11.4	Water Well Drilling		11-17
12.0	SAMPLING METHOD AND APPROACH			12-1
	12.1	Introduction		12-1
	12.2	Soil Sampling Procedures		12-1
		12.2.1	Nevsun Soil Sampling Procedures	12-1
		12.2.2	Mercier Soil and Auger	12-3
		12.2.3	Termite Mound Sampling	12-3
		12.2.4	Pit Sampling	12-3
	12.3	Rock Chip Sampling Procedures		12-4
	12.4	pH Survey Procedures		12-4
	12.5	Stream Sediment Sampling Procedures		12-4
	12.6	Trench Sampling Procedures		12-5
	12.7	Drill Core Sampling Procedures		12-5
	12.8	Reverse Circulation Drill Sampling Procedures		12-9
13.0	SAMPLE PREPARATION, ANALYSES AND SECURITY			13-1
	13.1	Introduction		13-1
	13.2	Sample Preparation for Soils and Sediment		13-2
		13.2.1	Nevsun – ALS Chemex	13-2
		13.2.2	Mercier – Horn of Africa	13-2
		13.2.3	Stream Sediment Sample Preparation	13-2
		13.2.4	Soil and Auger Sample Preparation	13-3
		13.2.5	Pit Sample Preparation	13-3

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	TOC iii
	October 2004

		13.2.6	Termite Mound Sample Preparation	13-4
	13.3	Sample Preparation of Drill Core and Rocks		13-4
		13.3.1	Horn of Africa Preparation Laboratory	13-4
		13.3.2	ALS Chemex	13-5
		13.3.3	Nevsun Sample Preparation Laboratory	13-7
	13.4	Sample Preparation of RC Chips		13-9
	13.5	Analyses		13-9
		13.5.1	Genalysis Laboratory Services	13-9
		13.5.2	ALS Chemex	13-10
	13.6	Nevsun Quality Assurance/Quality Control Program		13-11
	13.7	Security		13-17
14.0	DATA VERIFICATION			14-1
	14.1	Data Verification by Nevsun		14-1
	14.2	Data Verification by AMEC		14-1
	14.3	AMEC Quality Control Checks		14-3
	14.4	AMEC Independent Sampling		14-4
		14.4.1	Quartered Core	14-5
		14.4.2	Sub-sampling of Reject Material	14-5
		14.4.3	Splits Versus Original Sample	14-8
		14.4.4	Standards	14-9
		14.4.5	Blanks	14-10
		14.4.6	Bulk Density Checks	14-10
15.0	ADJACENT PROPERTIES			15-1
16.0	MINERAL PROCESSING AND METALLURGICAL TESTING			16-1
	16.1	Cyanidation		16-1
	16.2	Flotation		16-2
	16.3	Conclusion		16-2
17.0	MINERAL RESOURCE AND MINERAL RESERVE ESTIMATES			17-1
	17.1	Current Work		17-1
	17.2	Summary, Conclusions and Recommendations		17-1
		17.2.1	Resource Database and Geological Models	17-1
		17.2.2	Summary of Ore Controls	17-1
		17.2.3	Grade Capping	17-2
		17.2.4	Variography	17-3
		17.2.5	Block Model Validation	17-3
		17.2.6	Resource Classification	17-4
		17.2.7	Resource Summaries	17-5
	17.3	Geological Models and Creation of Vulcan Databases		17-6

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	TOC iv
	October 2004

		17.3.1	Introduction	17-6
		17.3.2	Geological Models and Assay Database	17-6
		17.3.3	Reduction of High Grades (Capping)	17-7
	17.4	Compositing and Exploratory Data Analysis		17-9
		17.4.1	Introduction	17-9
		17.4.2	Choice of Composite Length	17-9
		17.4.3	Statistical Distributions	17-10
		17.4.4	Contact Analysis	17-10
		17.4.5	Summary of Ore Controls	17-13
	17.5	Variography		17-14
		17.5.1	Introduction	17-14
		17.5.2	Variogram Models	17-15
	17.6	Estimation Plan		17-17
		17.6.1	Block Model Setup	17-17
		17.6.2	Estimation Plan Parameters	17-18
		17.6.3	Grade Estimation	17-19
		17.6.4	Implementation of Grade Capping Strategy	17-19
		17.6.5	Discussion	17-19
	17.7	Bulk Density		17-20
	17.8	Model Validation		17-22
		17.8.1	Introduction	17-22
		17.8.2	Visual Inspection	17-22
		17.8.3	Comparison to Nearest Neighbour Estimation	17-22
		17.8.4	Change of Support Check	17-23
	17.9	Resource Classification and Summaries		17-28
		17.9.1	Introduction	17-28
		17.9.2	Confidence Intervals for Grade Estimation	17-28
		17.9.3	Discussion	17-29
		17.9.4	Resource Summaries	17-31
18.0	OTHER RELEVANT DATA AND INFORMATION			18-1
19.0	REQUIREMENTS FOR TECHNICAL REPORTS ON PRODUCTION AND DEVELOPMENT PROPERTIES			19-1
20.0	CONCLUSIONS AND RECOMMENDATIONS			20-1
	20.1	Conclusions		20-1
	20.2	Recommendations		20-2
		20.2.1	Phase I Work Program	20-4
		20.2.2	Phase II Work Program	20-6
21.0	REFERENCES			21-1

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	TOC v
	October 2004

TABLES
Table 1-1	Summary of the Bisha Resource Estimate (Brisebois, 2004)	1-5
Table 1-2	Summary of the Oxide and Supergene Zone Resource Estimates (Brisebois, 2004)	1-6
Table 1-3	Summary of the Primary and Primary Zn Resource Estimate (Brisebois, 2004)	1-6
Table 4-1	UTM Coordinates of the Bisha Exploration License (UTM Zone 37)	4-1
Table 4-2	Bisha Prospecting and Exploration Licenses	4-3
Table 5-1	Distances by Road to the Bisha Exploration License	5-1
Table 6-1	General History of Bisha Property	6-1
Table 6-2	Phelps Dodge Corp. Grab Samples 1999	6-2
Table 6-3	2002 Drill Program Summary of Significant Assay Intervals	6-3
Table 7-1	Asmara Area Base Metal Prospects and Deposits	7-7
Table 7-2	Mineral Deposits in Eritrea, Sudan, Ethiopia and Western Saudi Arabia	7-9
Table 10-1	Summary of Work Complete	10-2
Table 10-2	Significant Assay Intervals from 2002 Drill Program	10-3
Table 10-3	Control Points in WGS84 (Geographic Coordinates)	10-4
Table 10-4	Summary of Geological Mapping on Bisha Property	10-4
Table 11-1	Drill Hole Summary by Year and Type	11-1
Table 11-2	Drill Program Survey Methods	11-4
Table 11-3	Recovery by Drill Program	11-9
Table 11-4	Average Recovery for Domain	11-10
Table 11-5	Nevsun Bulk Density Classification by Geological Domain for 2003	11-11
Table 11-6	Bulk Density Samples in Domain and Considered for 2004 Resource Estimate	11-13
Table 11-7	Mineralized Intervals for Each Geological Domain	11-14
Table 11-8	Distribution of Sample Intervals by Domain for each Hole Type	11-16
Table 11-9	Mineralized Intervals for Each Geological Domain	11-16
Table 11-10	Summary of Water Well Locations	11-17
Table 12-1	Total Line Kilometres of Soil Sampling on Bisha	12-2
Table 12-2	Summary of Trench Locations	12-6
Table 12-3	Summary of Drill Techniques Used	12-7
Table 12-4	Summary Statistics of Sample Lengths Grouped by Rock Type	12-8
Table 13-1	Summary of Standards Used on the Drill Programs	13-11
Table 13-2	Summary of Standard for Core and RC Sampling	13-13
Table 13-3	Summary of Blanks Greater than the 3x Detection Level	13-13
Table 14-1	Sieve Checks for Samples from 2004 Program	14-3
Table 14-2	Sieve Checks for Samples from Pre-2004 Programs	14-4
Table 14-3	Independent and QAQC Sampling	14-5
Table 14-5	Blank Samples Submitted with AMEC Samples	14-10

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	TOC vi
	October 2004

Table 14-6	Bulk Density Measurements using Wax and Non-Wax Immersion Methods	14-11
Table 14-7	Bulk Density Measurements using Wax Immersion Methods	14-11
Table 14-7	Bulk Density Measurements using Wax Immersion Methods	14-12
Table 17-1	Summary of the Bisha Mineral Resource Estimate (Brisebois, 2004)	17-5
Table 17-2	Summary of the Oxide and Supergene Zone Resource Estimates (Brisebois, 2004)	17-5
Table 17-3	Summary of the Primary and Primary Zn Resource Estimate (Brisebois, 2004)	17-6
Table 17-4	Metal-at-Risk Results from Capping Simulations	17-8
Table 17-5	Summary of Contact Analysis Results	17-13
Table 17-6	Estimation Domains	17-14
Table 17-7	Variogram Models	17-16
Table 17-8	Block Model Parameters	17-17
Table 17-9	Search Ellipse Parameters	17-19
Table 17-10	Bulk Density Values by Domain and Rock Type within Each Domain	17-21
Table 17-11	Comparison of Kriged Versus Nearest Neighbour – Indicated Blocks	17-22
Table 17-12	Change of Support Analysis	17-23
Table 17-13	90% Confidence Intervals for Metal Estimation in Various Domains	17-30
Table 17-14	Indicated and Inferred Resources by Zn Cut-offs	17-31
Table 17-15	Indicated and Inferred Resources by Au Cut-offs	17-32
Table 17-16	Indicated and Inferred Resources by Cu Cut-offs	17-33
Table 20-1	Phase I Work Plan and Budget	20-6
FIGURES
Figure 4-1	Location Map of Eritrea	4-2
Figure 4-2	Summary of Prospecting and Exploration Licenses in Eritrea	4-3
Figure 4-3	Location of the Bisha Exploration License	4-4
Figure 5-1	Property Access and Topography	5-2
Figure 7-1	Geology of the Arabian-Nubian Shield in the Red Sea Region	7-1
Figure 7-2	Geological Terrane Map of Eritrea and Nevsun Exploration Licenses	7-2
Figure 7-3	General Geology Map of Eritrea	7-4
Figure 7-4	Structural Interpretation of Western Eritrea	7-6
Figure 7-5	Stratigraphic Section	7-11
Figure 7-6	Property-Scale Geology Map	7-14
Figure 7-7	Deposit-Scale Geology Map	7-15
Figure 8-1	Kuroko Style VMS Deposit Model	8-2
Figure 8-2	Kuroko Style VMS Grade and Tonnage Model (Singer and Mosier, 1986)	8-3
Figure 8-3	Bisha Bimodal Siliciclastic VMS Model Schematic	8-4
Figure 9-1	Isometric View of the Bisha Deposit Facing West	9-1
Figure 9-2	Drill Hole Location and Bisha Main Zone Outline	9-3

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	TOC vii
	October 2004

Figure 10-1	Bisha Exploration License Geological Mapping Location Map	10-5
Figure 10-2	1998 and 2003 Stream Sediment Survey Results	10-7
Figure 10-3	Soil Sample Results Over the Bisha Gossan Zone	10-8
Figure 10-4	Geochemical Sampling Areas on the Bisha Property	10-10
Figure 10-5	Ground Geophysical Survey Compilation Map	10-12
Figure 10-6	Gravity Coverage on Bisha Property	10-14
Figure 11-1	Drill Hole Location Map	11-2
Figure 11-2	Average Azimuth and Dip Change Compared with Collar	11-6
Figure 11-3	Change in Azimuth with Dip	11-7
Figure 13-1	Particle Size Distribution Graph for Crushing	13-6
Figure 13-2	Particle Size Distribution Graph for Pulverizing	13-6
Figure 13-3	Graphs for Twin Samples for Gold, Silver, Copper, Lead and Zinc	13-15
Figure 13-4	Graphs for Coarse Preparation Duplicates for Gold, Silver, Copper, Lead and Zinc	13-16
Figure 14-1	Original versus Quarter Core Sample Pairs	14-6
Figure 14-2	Original versus Reject Sample Pairs	14-7
Figure 14-3	Original versus Last Split Sample Pairs	14-8
Figure 17-1	3 m Composite Statistical Comparison by Lithologic Domain – Zn (%)	17-10
Figure 17-2	3 m Composite Statistical Comparison by Lithologic Domain – Cu (%)	17-11
Figure 17-3	3 m Composite Statistical Comparison by Lithologic Domain – Au (g/t)	17-11
Figure 17-4	3 m Composite Statistical Comparison by Lithologic Domain – Ag (g/t)	17-12
Figure 17-5	3 m Composite Statistical Comparison by Lithologic Domain – Pb (%)	17-12
Figure 17-6	Orientation Views of the Block Model	17-18
Figure 17-7	Grade Profiles Comparing Kriged to Nearest Neighbour Block
	Estimates in the Indicated Blocks (Au and Estimated Tonnage)	17-24
Figure 17-8	Grade Profiles Comparing Kriged to Nearest Neighbour Block
	Estimates in the Indicated Blocks (Ag and Zn)	17-25
Figure 17-9	Grade Profiles Comparing Kriged to Nearest Neighbour Block
	Estimates in the Indicated Blocks (Pb and Cu)	17-26
APPENDICES
A	Photographs
B	AMEC Independent Sampling Details and Results
C	Drill Hole Summary
D	Nevsun and ALS Chemex Sampling and Analytical Protocols
E	Mineralogical and Metallurgical Studies
F	Land Tenure and Surface Rights Documentation
G	Geological Model and Resource Model Supporting Information

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	TOC viii
	October 2004

1.0

SUMMARY

1.1

Introduction

Nevsun Resources (Eritrea) Ltd., a subsidiary of Nevsun Resources Ltd. (Nevsun) requested AMEC (Perú) S.A. (AMEC) to provide an independent Qualified Person’s Review and Technical Report of the Bisha Property (the Property), which is located in the District of Gash-Barka, Eritrea. Doug Reddy, P.Geo., a member of the APEGBC and AUSIMM and an employee of AMEC (Lima, Peru office), and Ken Brisebois, P.Eng., Consulting Engineer with AMEC (Phoenix office) served as the Qualified Persons responsible for the preparation of the Technical Report as defined in National Instrument 43-101, Standards of Disclosure for Mineral Projects and in compliance with Form 43-101F1 (the Technical Report). Ken Brisebois prepared the resource estimate.

The scope of work entailed review of pertinent geological, geophysical, and other data in sufficient detail to prepare a mineral resource estimate and the Technical Report. Only limited early stage metallurgical testwork was available and as such AMEC has summarized that information and completed a brief review by a process engineer (Lynton Gormely, P.Eng., Process Engineer, AMEC Vancouver office). AMEC understands that the Technical Report will be submitted by Nevsun to the Toronto Stock Exchange (TSX).

Nevsun holds a 90% interest in the 322 km2 Bisha Exploration License through its wholly owned subsidiary, Nevsun Resources (Eritrea) Ltd. The State of Eritrea holds the remaining 10% interest. Nevsun has temporarily suspended fieldwork in response to a stop work order from the Government of Eritrea issued to all exploration companies. The circumstances of the order are not public.

Initial prospecting by Amanuel Woldu (Ophir Ventures) in 1996 included sampling of a gossanous area that is now recognized as the Bisha Main Zone. Nevsun acquired the property in 1998 following a site visit by Bill Nielsen. Subsequent fieldwork included mapping (1:50,000 scale) and stream sediment sampling. In 1999, detailed mapping (1:5,000 scale), soil sampling, and geophysics (HLEM, magnetometer) was completed. Work was suspended during a two-and-a-half year border war with Ethiopia that erupted in 1998 and ended in December 2000. Drilling of 6 holes in late 2002 intercepted significant intervals of precious and base metal mineralization at the Bisha Main Zone. Nevsun mounted a major exploration effort in February 2003, which included trenching, geophysics (ground and airborne surveys) and drilling to define the mineralization. Further exploration and drilling programs were conducted in September 2003 and January 2004.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-1
	October 2004

Nevsun has completed a total of 50,715.76 m of drilling in 352 holes. Of this, 48,309.66 m was core drilling in 310 holes; 1,808.4 m was completed in 33 RC holes; and 591.70 m was in 9 combination holes, which had RC at the top of each hole and core drilling in the bottom.

1.2

Geology and Mineralization

Bisha is a precious and base metal-rich volcanogenic massive sulphide (VMS) deposit. Pertinent deposit model types would be Noranda/Kuroko (Franklin et. al., 1981) or bimodal-siliciclastic VMS deposits (Barrie, 2004).

Four principal zones of mineralization within the Bisha Main Zone include: (1) a near-surface oxide/gossan; (2) a horizon that has been subjected to extreme acidification (acidified); (3) a supergene copper-enriched horizon; and (4) a primary massive sulphide horizon.

Characteristics of the host units to the Main Zone mineralization include:

• Precious metal (Au, Ag) and base metal rich (Cu, Zn, Pb) massive sulphide lenses hosted by a bimodal sequence of weakly stratified, predominantly tuffaceous metavolcanic rocks (Nacfa Terrane greenstone belt).
  

• Host rocks are felsic lithologies (variably altered felsic lapilli and lapilli ash tuffs, crystal tuffs and minor felsic dykes), which also form the hanging wall stratigraphy and predominate overall.
  

• Sub-alkaline (Greig, 2004) geochemistry of the volcanic rocks.
  

The Main Zone is a 1.2 km long long, narrow massive sulphide lens oriented north-south. The true thickness of the lens is variable from 0 to 70 m. The deposit is deformed and exhibits thickening at the fold hinge and limb attenuation, which distorts original dimensions. Drill hole intersections encountered mineralization to a depth of 380 m but portions of the deposit only extend to depths of 70 m.

A fault with a northwest strike is interpreted to have displaced the Main Zone upwards (displaced northeast side up) at the northern end (at 1716000 N; Greig, 2004). Another northwest trending fault has been postulated to occur near the centre of the lens (at 1715400 N) where the western lens (or limb) disappears from the sections possibly due to vertical displacement of the deposit. The understanding of the structural history of the deposit is not complete and differing interpretations are possible.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-2
	October 2004

The south end of the Bisha Main Zone plunges very rapidly and drilling has been completed without success for 50 m south of the last mineralized intercept. The north end of the Main Zone appears to be abruptly terminated due the exposure and erosion of the keel of the Bisha Syncline. The deposit remains open down dip in several portions of the deposit. Extensions at depth would add primary sulphide mineralization.

Metal zoning within the massive sulphide appears to indicate an upward transition from Cu-rich to Zn-rich to barren pyrite and this confirms the interpretation that the sequence is right-way-up.

1.3

Data, QAQC, EDA and Verification Sampling

Nevsun conducted a QAQC program to monitor the accuracy and precision of the assays and ensure that sampling, preparation and analytical protocols were being maintained. The results were reviewed by AMEC and are acceptable.

During the site visit AMEC collected independent quarter core and check samples. The check samples include rejects of crushed reject material from the storage facility for comparison against original sample assays. QAQC checks were also submitted including standards, blanks, and preparation duplicates (first split and last split). The samples were renumbered, randomized and submitted “blind”. The samples remained in AMEC custody until delivered to the Ministry of Mines in Asmara for customs inspection and shipping. The samples were placed in the care of the Ministry due to restrictions on export of rock samples. Results from the check sampling were acceptable and did not identify any problems or concerns.

AMEC considers the practices used by Nevsun to collect the data to be in accordance with standard industry practices. AMEC completed a thorough review and verified the database used for resource estimation. Problems with the database were resolved as they were identified and the final database was suitable for use in resource estimation.

1.4

Metallurgical Testwork

Metallurgical testwork included cyanidation of oxide material (2 samples) and flotation tests of both supergene copper enriched material (2 samples) and primary massive sulphide mineralization (4 samples) from rejects of crushed samples from drill cores. All testwork was completed by Process Reseach Associates (PRA) based in Vancouver, Canada.

The samples submitted for cyanidation were core of the high-grade oxide intervals and had gold grades of 23.3 and 10.8 g/t Au respectively. Diagnostic leach tests suggest that 90 to 95% of the gold is directly cyanide soluble. In 72 hours cyanidation (PRA test C3), the 23.3 g/t Au sample yielded 96.1% gold recovery, or a residue of 0.91 g/t Au. Silver in the

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-3
	October 2004

residue was 10.7 g/t. In 72 hours cyanidation (PRA test C4), the 10.8 g/t Au sample yielded 82.1% recovery, or a residue of 1.97 g/t Au. Cyanide consumptions for the two tests were low at 0.17 and 0.33 kg/t, respectively. Lime consumptions were also low to moderate at 1.78 and 3.73 kg/t, respectively.

Possible options to improve recoveries include: ultra fine grinding, intensive cyanidation or a combination of these methods. Gold is reportedly micron size and therefore gravity processing of the material may not be feasible.

The core samples of supergene copper mineralization that were submitted for flotation tests had grades of 8% Cu and 3.8% Cu, respectively. The concentrate grades that were achieved were not upgraded significantly from the (possible) head grades of the samples, as estimated above. Flotation may not have been successful in upgrading the samples due to the high pyrite content (~95%) in the products. The mineralogical report for the samples identified high amounts of pyrite.

Four additional samples collected from the massive sulphide component of the resource were tested for upgrading by flotation. Copper recoveries of the samples ranged from 79 to 94%, however gold and silver recoveries for these samples through flotation were poor at 38 to 51% Au recovery and 53 to 77% Ag recovery.

The deposit mineralogy is polymetallic and will require a significant testwork program to identify an optimum process flowsheet and to confirm metal products, grades and recoveries. Process design will likely include several process recovery circuits – i.e. sulphide flotation with several circuits (i.e. Cu and Zn banks) followed by cyanide leaching of flotation tails.

1.5

Resource Estimation

Nevsun and AMEC completed geological modelling. The geological model was developed using sectional spacing of 25 m for the basic interpretations that were subsequently rationalized in plan. Solid (wireframe) models were created for each principal mineralized or geological domain and became the basis for coding the block model.

AMEC prepared the mineral resource estimate using industry standard methodologies conforming to the requirements set out in National Instrument 43-101. The modelling was carried out using Vulcan 3D Software by Ken Brisebois, P.Eng.

The blocks grades were estimated within domains based on interpretation of geological parameters logged in drill holes. Grades were estimated using Ordinary Kriging utilizing two passes and search neighbourhoods conforming to the geological trend.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-4
	October 2004

Capping of extreme grades was used to remove metal at risk as derived from analyses of the assay distributions. Calibrated in the Indicated blocks of the model, approximately 7.5% of the Au, 2.5% of the Zn, 13.5% of the Ag and 3.6% of the Cu was removed by the capping process.

Bulk density was estimated by using the average of the measurements from within the modeled zones of mineralization. Extreme or potentially erroneous values were scrutinized or removed from the dataset. A full discussion of bulk density values used for the resource estimate is provided in Section 11.2.7.

Variography and confidence limits analyses for grade estimation were used in conjunction with confidence in geological modelling and database integrity to develop resource classification criteria. Indicated resources were defined by a nominal 25 m spaced drill sampling grid. Inferred resources were defined to be the remainder of the material within 50 m of drilling. Both the Indicated and Inferred material were restricted to the geologically interpreted mineralization domains.

A total of 22.75 Mt was reported as Indicated mineral resources, and an additional 5.85 Mt was also reported as Inferred mineral resources (see Table 1-1). Table 1-2 summarizes the mineral resources of the Oxide and Supergene Zone and Table 1-3 summarizes the mineral resources of the Primary Zone.

Table 1-1
Summary of the Bisha Resource Estimate (Brisebois, 2004)

Category	Zone	Cut-off	Tonnes (000’s)	Au g/t	Ag g/t	Cu %	Zn %
Indicated	Oxides	0.5g/t Au	4,984.1	6.51	30.0	0.10	0.08
	Supergene Cu	0.5% Cu	7,644.8	0.46	35.56	3.47	0.87
	Primary	2.0% Zn	1,711.5	0.74	29.59	0.97	3.07
	Primary Zn	2.0% Zn	8,413.3	0.76	58.27	1.12	9.04
	Total tonnes		22,753.7
Inferred	Oxides	0.5g/t Au	122.0	3.34	18.2	0.12	0.07
	Supergene Cu	0.5% Cu	185.6	0.09	30.14	3.26	1.04
	Primary	2.0% Zn	392.0	0.75	35.20	1.24	3.03
	Primary Zn	2.0% Zn	5,150.9	0.70	59.67	0.84	8.28
	Total tonnes		5,850.5

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-5
	October 2004

Table 1-2
Summary of the Oxide and Supergene Zone Resource Estimates (Brisebois, 2004)

Zone	Category	Domain	Cut-off Au g/t	Tonnes (000’s)	Au g/t	Ag g/t	Cu %	Zn %	Oz millions
Oxide	Indicated	Fe Oxide	0.5	3,653.4	6.65	20.32	0.10	0.08	0.781
Oxide	Indicated	Acidified	0.5	709.6	8.35	108.05	0.09	0.03	0.190
Oxide	Indicated	Breccia	0.5	621.1	3.62	9.39	0.09	0.07	0.072
Oxide	Indicated	Fe Oxide	1.0	3,469.5	6.97	20.79	0.10	0.08	0.778
Oxide	Indicated	Acidified	1.0	683.7	8.63	109.88	0.10	0.03	0.189
Oxide	Indicated	Breccia	1.0	519.1	4.19	10.65	0.09	0.07	0.069
Zone	Category		Cut-off % Cu	Tonnes (000’s)	Au g/t	Ag g/t	Cu %	Zn %	lbs Cu millions
Supergene	Indicated		0.25	8,105.2	0.44	34.5	3.30	0.86	589
Supergene	Indicated		0.5	7,644.8	0.46	35.6	3.47	0.87	585
Supergene	Indicated		1.0	6,453.1	0.50	38.7	3.97	0.91	564

Table 1-3
Summary of the Primary and Primary Zn Resource Estimate (Brisebois, 2004)

Zone	Category	Cut-off Zn %	Tonnes (000’s)	Au g/t	Ag g/t	Cu %	Zn %	Contained Zn Millions lb	Contained Cu Millions lb
Primary	Indicated	2.0	1,711.5	0.74	29.6	0.97	3.07	115	37
Primary Zn	Indicated	2.0	8,413.3	0.76	58.3	1.12	9.04	1,680	207.7
Primary	Inferred	2.0	392.0	0.75	35.2	1.24	3.03
Primary Zn	Inferred	2.0	5,150.9	0.70	59.7	0.84	8.28

The Bisha Main Zone is the principal mineral resource on the Bisha Property and the focus of exploration work and current studies. Additional drilling will be required to advance the resource estimate to a feasibility study level. In Section 1.7 AMEC provides recommendations to advance the project towards engineering studies.

Recent structural and geological studies confirm the geologic model and support the opportunities for further exploration success. At the south end of the deposit the massive sulphides plunge off steeply but significant widths of mineralization containing appreciable zinc values are present at depth.

A stringer sulphide zone is located in the footwall of the primary sulphide zone. The stringer zone contains significant base and precious metal values. Nevsun is currently modelling this potential resource.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-6
	October 2004

A widespread zone of mineralization consisting of disseminated copper sulphides in oxidized rocks has also been identified. Holes such as B-111 that encountered 39 m grading 1.08% Cu and B-87 that intersected 31.5 m assaying 1.44% Cu are examples of this disseminated style of mineralization. The potential resources available in this area are being assessed.

Other targets on the Bisha Property include the Northwest Zone, Harena, and the NW Barite Showing areas.

The Northwest Zone is interpreted to be a structural repetition of the stratigraphy hosting the Bisha Main Zone mineralization. Mapping, geochemistry, and geophysics all support this as a good exploration target with clear anomalies. Drilling intersected mineralization in 8 of the 14 holes and one of the longest intervals is from hole B-066 with a 47.5 m core length interval averaging 1.32 g/t Au, 14.96 g/t Ag, 1.52% Cu, 0.01% Pb, and 0.04% Zn (pers. comm. Nielsen, 2004). The true width of the interval is not yet known because the interpretations are preliminary. This zone warrants additional drilling.

The Harena Area is 9 km southwest of the Bisha Main Zone and has a gossan with associated geochemical and geophysical anomalies. The target warrants additional exploration including drilling.

The NW Barite Showing has a geochemical and geophysical anomaly. The target warrants additional exploration including drilling.

1.6

Recommendations

During the site visit and review AMEC has made several recommendations for improvements to the work being conducted on the Bisha Property. Nevsun has already addressed many of the recommendations presented by AMEC during this study (i.e. corrections to database, change in magnetic declination, etc.). Other items that should be addressed are as follows:

• Implement a double data entry system or a data entry system with some form of validation of codes. Entry of data could be direct to MS Access or a relational database with filters, limits, and data integrity checks.
  

• Logging code system should be simplified, standardized and a set of equivalent codes should be prepared to equate the logging to surface mapping.
  

• Include a commercial blank sample within the QA/QC program. If the coarse blank material is also continued then care should be exercised during collection of the material.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-7
	October 2004

• Assay a series of pulp samples (check assays) at an external laboratory (approximately 5% is a normal recommended number of check assays). Also, continue to submit 1 in 20 samples to a second independent laboratory other than ALS Chemex.
  

• Check the relative coverage of the bulk density measurements. If sufficient samples can be collected from earlier drilling or current drilling then assess the possibility of developing a density model for resource estimation.
  

AMEC considers that further drilling and engineering studies are warranted on the Bisha Property to advance the Bisha Deposit towards feasibility studies. The recommended activities include:

• Geotechnical assessment of the potential open pit parameters. The available geotechnical data, including oriented core should be reviewed and modelled. Geotechnical data collection involving drilling into the walls of the potential open pit will be required. A geotechnical engineer should provide the coordinates for geotechnical drilling and a series of procedures for logging and sampling of geotechnical information.
  

• Metallurgical samples should be collected and tested. The suite of samples must be representative of the mineralized domains and consider the specialized testwork required for each area of the deposit. Sample collection will require large diameter core (PQ - 85 mm diameter) for metallurgical samples. A metallurgical testwork program should be designed to assess process options and provide sufficient information for preliminary and/or feasibility studies. A reputable metallurgical testing facility needs to be retained to complete and report on this work.
  

• Baseline environmental studies, including social and archaeological studies (underway), in preparation for an Environmental Impact Assessment (EIA).
  

• Socio-economic assessment (underway).
  

• Hydrological studies (underway).
  

• Tailings containment system design (underway).
  

• Waste and tailings acid generation assessment (underway).
  

• Complete drilling to infill the near-surface resources of the Bisha Main Deposit. Nevsun estimates that this will involve approximately 1,500 m of core drilling. To improve drill recoveries in the gossan area the holes should start with PQ-sized core and subsequently reduce to HQ and NQ as the competency of the rock increases with depth and the effects of surface weathering and oxidation diminish.
  

• Complete closer-spaced drilling on sections 12.5 m apart to provide improved confidence and advance a significant portion of the resources to the Measured category.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-8
	October 2004

• Review the model and prepare an updated resource estimate including the mineralization in the stockwork and disseminated zones.
  

• Improve the roads and basic infrastructure in the immediate Bisha area.
  

All drilling and exploration work should include the QAQC programs and practices that are already in place. Nevsun and AMEC also consider that the following exploration activities should be completed:

• Deeper drilling on the Bisha Main Zone to determine the overall extent of the VMS deposit, i.e.:
  

-

Down-plunge on the south end of the Main Zone.

-

Along trend of mineralized horizon at north end of the Main Zone.

-

Down-dip of eastern limb.

• Drill the Bisha Northwest VMS Zone.
  

• Drill the prime exploration targets such as the coincident gravity, HLEM and soil anomalies defined on the Harena Area (SW grid).
  

• Model the gravity targets south of line 1715000N using available specific gravity data from various core intervals to determine the depth of any potential massive sulphide mineralization that may be present in this area.
  

• Conduct exploration and possible drilling in the NW Barite Showing area.
  

1.6.1

Phase I Work Program

Nevsun has provided a budget to address Phase I the activities described above which include:

• Geotechnical testwork.
  

• Metallurgical testwork.
  

• Environmental assessment (ongoing studies).
  

• Archaeological assessment (ongoing studies).
  

• Tailings containment system design (ongoing studies).
  

• Waste and tailings acid generation assessment (ongoing studies).
  

• Socio-economic assessment (ongoing studies).
  

• Infill drilling near-surface around the gossan.
  

• Drilling around the Bisha Main Zone.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-9
	October 2004

• Update resource estimates.
  

• Improve roads.
  

• Exploration and drilling additional of targets listed above.
  

The total budget for Phase I activities is US$2.76 M. Exploration activities are approximately US$2.01 M (72%), engineering studies and testwork comprises approximately US$0.435 (16%), and camp operation and support is approximately US$ 0.313 M (11%) of the budget. This work should commence once the stop work order is lifted.

1.6.2

Phase II Work Program

Furthermore AMEC considers that the Bisha Project should be advanced towards completion of a feasibility study.

AMEC recommends that the studies competed in Phase I will be used to develop a Scoping Study to identify the overall conceptual project scope and address: geology, mining, process, ancillary facilities, infrastructure, environmental, opportunities, risks, capital costs, operating costs and financial analysis. Costs to complete a Scoping Study could be US$0.100 to 0.200 M including the additional investigations required for preliminary mine plan, flowsheet, cost estimates etc. and the number of options considered.

The Scoping Study may show that additional studies are required or it may conclude that the next step is to commence a detailed Feasibility Study.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 1-10
	October 2004

2.0

INTRODUCTION AND TERMS OF REFERENCE

2.1

Introduction

Nevsun Resources (Eritrea) Ltd., a subsidiary of Nevsun Resources Limited (Nevsun) commissioned AMEC (Perú) S.A. (AMEC) to provide an independent Qualified Person’s Review and Technical Report of the Bisha Property (the Property). Doug Reddy, P.Geo., Principal Geologist (AMEC Lima, Peru office) and Ken Brisebois, P.Eng., Consulting Engineer (AMEC Phoenix office) served as the Qualified Persons responsible for the preparation of the Technical Report as defined in National Instrument 43-101, Standards of Disclosure for Mineral Projects, and in compliance with Form 43-101F1 (the Technical Report). Mr. Reddy has more than 18 years of experience in the mining industry, including substantial experience with volcanogenic massive sulphide deposits and in precious and base metal exploration. Ken Brisebois prepared the mineral resource estimate.

Information and data for AMEC’s review and report were obtained from Nevsun, Taiga Consultants Ltd. (Taiga), and various independent consultants (C. Greig., T. Barrie, P. Andersen, D. Outtara). Mr. Reddy, completed a 5 -day site visit between May 28 and June 1, 2004 and during this time visited the camp, offices, core logging and storage facilities, sample preparation facility and the main prospect areas on the Property.

Assistance during the site visit and subsequent studies related to this report was provided by:

• John Clark, President & Chief Executive Officer, Nevsun Resources Ltd.
  

• Bill Nielsen, V.P. of Exploration, Nevsun Resources Ltd. (Nevsun’s Q.P. for Exploration Program Design and Operation).
  

• Greg Davis, Site Manager, Database Manager, Nevsun Resources Ltd.
  

• Scott Ansell, Program Manager and Senior Geological Technician, Nevsun Resources Ltd.
  

• David Daoud, Site Geologist, Nevsun Resources Ltd.
  

• Tony Odametey, Geotechnical Geologist, Nevsun Resources Ltd.
  

• Amanuel Woldu, Eritrean Country Manager, Nevsun Resources (Eritrea) Ltd.
  

• Craig Scherba, Geologist and Database Manager, Taiga Consultants Ltd.
  

• Claude Assaunt, Geologist and Project Manager, Taiga Consultants Ltd.
  

• Bob Nichol, Geologist, Taiga Consultants Ltd.
  

• Robin Chisholm, Geologist, Taiga Consultants Ltd.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 2-1
	October 2004

• Charlie Greig, Independent Geological Consultant.
  

• Tucker Barrie, Independent Geological Consultant.
  

• Paul Anderson, Independent Geological Consultant.
  

• Daouda Ouattara, Independent Geological Consultant.
  

2.2

Terms of Reference

AMEC is not an associate or affiliate of Nevsun, or of any associated company. AMEC’s fee for this technical report is not dependent in whole or in part on any prior or future engagement or understanding resulting from the conclusions of this report. This fee is in accordance with standard industry fees for work of this nature, and AMEC’s previously provided estimate is based solely on the approximate time needed to assess the various data and reach the appropriate conclusions.

In preparing this report, AMEC relied on geological reports and maps, miscellaneous technical papers listed in the References section at the conclusion of this report and AMEC’s experience on similar deposit types. During AMEC’s recent site visit, the surface exposures of the key exploration targets described in this report were visited and core was reviewed. AMEC collected a total of 172 samples and submitted them to ALS Chemex in Vancouver, Canada for preparation and analyses.

This report is based on information known to AMEC as of 1 October, 2004.

All measurement units used in this report are metric, and currency is expressed in US dollars unless stated otherwise. The currency used in Eritrea is the Nacfa. The exchange rate as of 1 October, 2004 is US $1.00, equal to approximately 14 Nacfa.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 2-2
	October 2004

3.0

DISCLAIMER

AMEC has not reviewed the land tenure, nor independently verified the legal status or ownership of the properties or underlying option and/or joint venture agreements. The results and opinions expressed in this report are based on AMEC’s field observations and the geological and technical data listed in the References (Section 21.0). While AMEC has carefully reviewed all of the information provided by Nevsun, and believes the information to be reliable, AMEC has not conducted an in-depth independent investigation to verify its accuracy and completeness.

The results and opinions expressed in this report are conditional upon the aforementioned geological and legal information being current, accurate and complete as of the date of this report, and the understanding that no information has been withheld that would affect the conclusions made herein. AMEC reserves the right, but will not be obliged, to revise this report and conclusions if additional information becomes known to AMEC subsequent to the date of this report. AMEC does not assume responsibility for Nevsun’s actions in distributing this report.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 3-1
	October 2004

4.0

PROPERTY DESCRIPTION AND LOCATION

4.1

Location

The Bisha Property consists of an exploration license located approximately 150 km west of Asmara (233 km by road), 43 km southwest of the regional town of Akurdat (Agordat on many maps), and 50 km north of Barentu, the regional or Zone Administration Centre of the Gash-Barka District (Figure 4-1), in Eritrea, East Africa.

The Property is at approximate latitude 15°24’N and longitude 37°30’SE (Figures 4-1 and 4-2). The UTM coordinates of the centre of the Property are 1,715,000 N and 340,000 E (UTM Zone 37).

4.2

Land Tenure

The Property is a single, contiguous exploration license¹ with dimensions of 14 km by 23 km and covering a total surface area of 322 km² (Figure 4-1). UTM coordinates of the Bisha Exploration License are listed in Table 4-1.

Table 4-1
UTM Coordinates of the Bisha Exploration License (UTM Zone 37)

Corner Point	Easting	Northing
A	331,000	1,718,000
B	354,000	1,718,000
C	354,000	1,704,000
D	331,000	1,704,000

Nevsun, through its wholly-owned Eritrea subsidiary Nevsun Resources (Eritrea) Ltd., holds a 90% interest in the property. The State of Eritrea holds the remaining 10%.

The original prospecting licenses were converted to exploration licenses each of which are granted as an agreement with the State of Eritrea (Proclamation No. 68/1995 A Proclamation to Promote the Development of Mineral Resources; pers. comm. Chisholm, 2004; Table 4-2, see Appendix F). Each exploration license is valid for three years and may be renewed twice for a term of one additional year of each renewal. The exploration license may be converted to a mining license if all obligations are met and an application with fees is submitted (see Section 4.3.6).

¹Nevsun also holds the Augaro and AK Exploration Licenses, in Eritrea but they are not discussed in this report.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-1
	October 2004

     Figure 4-1 Location Map of Eritrea

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-2
	October 2004

Figure 4-2
Summary of Prospecting and Exploration Licenses in Eritrea

The Bisha Exploration License includes the original Bisha Area Exploration License (obtained in 1999 and provided in Appendix F), the more recent Bisha Extension Area Exploration License (obtained in 2003) and the Okreb South and Okreb North Areas which were obtained in 1999 (Table 4-2; pers. comm. Chisholm, 2004). A description of the changes to land tenure is provided in Section 6.0 of this report.

Table 4-2
Bisha Prospecting and Exploration Licenses

Prospecting License Name	Date	Area (km2)	Exploration License Name	Date	Area (km2)	Expenditure Obligation (US$ ‘000)2			Annual Rental Fee (Nacfa)
						Year 1	Year 2	Year 3
Bisha Area	June 3, 1998	49	Bisha	June 22, 1999	49	190	1,000	500	9800
Okreb Area	June 3, 1998	98	Okreb North	June 2, 1999	49	135	500	250	9800
			Okreb South	June 21, 1999	49	135	500	250	9800
-	-	-	Bisha Extension Area	March 28, 2003	175	260	?	?	35,000
			Total		322	720	2,000	1000	64,400 Nacfa

² Due to the war with Ethiopia, the payments for the Bisha Exploration License, Okreb North and Okreb South were suspended until 2002.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-3
	October 2004

Figure 4-3
Location of the Bisha Exploration License

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-4
	October 2004

Annual license rental fees are 200 Nacfa per km², therefore an annual obligation of 64,400 Nacfa is due on the anniversary of the Bisha Exploration License.

The licenses have expenditure obligations during the first three years. Fee payments and expenditure obligations were suspended between 1999 and 2002 due to the border war with Ethiopia. At least US$3.720 M was required to have been spent on the licenses by 2004 and Nevsun considers that these commitments have been met during work programs in 2003 and 2004. Expenditure obligations for years 2 and 3 of the Bisha Exploration Area License were not stipulated in the license agreement.

AMEC relies on land tenure documentation supplied by Nevsun and Nevsun’s consultants and lawyers. AMEC was provided with a copy of the Bisha Area Exploration License Agreement (Appendix F) but not with the Okreb North, Okreb South or Bisha Extension Area License Agreements. An independent verification of title was not part of the scope of this study, nor has it been confirmed if additional pre-existing mining licences or concessions owned by other parties occur within any of the Property concessions, and which concessions would take precedence (Figure 4-2).

4.3

An Overview of Eritrea

4.3.1

Introduction

Eritrea was awarded to Ethiopia in 1952 by the United Nations as part of the establishment of a federation. Ten years later, Ethiopia's annexation of Eritrea as a province sparked a 30-year struggle for independence that ended in 1991 with Eritrean rebels defeating government forces. Independence was overwhelmingly approved in a 1993 Eritrean national referendum.

A two-and-a-half-year border war with Ethiopia erupted in 1998 and ended under UN auspices on 12 December, 2000. Eritrea currently hosts a UN peacekeeping operation that is monitoring a 25 km-wide Temporary Security Zone along the border with Ethiopia. An international commission, organized to resolve the border dispute, posted its findings in 2002 but the final demarcation of the border is on hold due to Ethiopian objections to the location of the border as presented.

4.3.2

Geography and Infrastructure

Eritrea is located above the Horn of Africa on the continent’s east coast, between Sudan to the north and west, and Ethiopia and Djibouti to the south. Eritrea has a 1,151 km coastline on the Red Sea, which separates the country from Saudi Arabia and Yemen (source: www.cia.gov/cia/publications/factbook/geos/er.html).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-5
	October 2004

Eritrea has an area of 124,320 km². The country is divided into three main geographical zones: (1) the fertile and intensively farmed mountainous central plateau that varies from 1,800 to 3,000 masl; (2) the eastern escarpment and coastal plain which are mainly desert; and (3) semi-arid western lowlands. There are over 350 islands located along the coast of Eritrea within the Red Sea and the Dahlak Archipelago (Figure 4-1). Eritrea has no year-round rivers.

The climate is temperate in the mountains and hot in the lowlands. Asmara, the capital, is located is about 2,300 m (7,500 ft.) above sea level. The maximum temperature is 26ºC (80ºF). The weather is usually sunny and dry, with the short or “belg” rains occurring between February to April and heavy or “meher” rains beginning in late June and ending in mid-September (source: www.state.gov/p/af/ci/er).

There is a good network of paved roads connecting Asmara with the major regional centres of Keren, Massawa, Adi Quala and Barentu. A paved road is under construction along the coast from Massawa to Assab (see Figure 4-1). Power generation from the Hirgigo diesel plant near Massawa supplies electrical power to Asmara and other major regional centres. Landline telephone service is available from larger towns and cellular service was recently announced in Asmara and surrounding towns; including Keren.

Comprehensive medical services are found in the larger towns with rudimentary medical clinics available in the smaller villages. Schools are located even in the smallest of villages.

4.3.3

Modern History

Eritrea was part of the first Ethiopian kingdom of Aksum until its decline in the 8th century. The area came under the control of the Ottoman Empire in the 16th century, and later of the Egyptians. Italy captured the coastal areas in 1885, and the Treaty of Uccialli (May 2, 1889) gave Italy sovereignty over part of Eritrea. The Italians named the colony after the Roman name for the Red Sea, Mare Erythraeum, and governed until World War II.

Britain invaded and took control of Eritrea in 1941 and later administered it as a UN Trust Territory until it became federated with Ethiopia on September 15, 1952. On November 14, 1962 Ethiopia established Eritrea as a province, which sparked a 30-year struggle for independence.

In 1991, the Ethiopian People's Revolutionary Democratic Front deposed Ethiopia’s hard-line communist dictator Mengistu. Without the Ethiopian troops to battle, the Eritrean People's Liberation Front (EPLF) was able to gain control of Asmara, the Eritrean capital, and form a provisional government. In 1993, a referendum on Eritrean independence was held, supported by the UN and the new Ethiopian government. Eritrean voters almost unanimously opted for an independent republic. Ethiopia subsequently recognized Eritrea's

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-6
	October 2004

sovereignty on May 3, 1993, and sought a new era of cooperation between the two countries. Independence of Eritrea left Ethiopia landlocked (Figure 4-1).

Following Eritrea's independence, Eritrea and Ethiopia disagreed on the demarcation of their mutual borders, and in May 1998 border clashes broke out. After an eight-month lull during which time both sides reinforced their 912 km common border, war broke out. About 80,000 people were killed and refugees fled to neighbouring countries. The war ended in a stalemate, and a formal peace agreement was signed in December 2000. The United Nations has supplied more than 4,000 troops to patrol a 25 km wide Temporary Security Zone between the two nations. An international boundary commission ruled on the disputed border between the two countries on April 13, 2002. Ethiopia disputed the new border demarcation as presented.

4.3.4

Demography and Government

Eritrea has an estimated population of approximately 4.45 million (July, 2004 estimate from CIA factbook, source: www.cia.gov/cia/publications/factbook/geos/er.html) however, this appears to be overestimated based on the number of inhabitants in the populated centres and more reasonably would be approximately 3.0 million. The majority of the people are disseminated throughout the countryside. Populated centres include the capital city of Asmara (est. pop. 435,000), Keren (57,000); Assab (28,000); Massawa (25,000); Afabet (25,000); Tessenie (25,000); Mendefera (25,000); Dekemhare (20,000); Adekeieh (15,000); Barentu (15,000); and Ghinda (15,000).

The population is composed of the following ethnic groups: Tigrinya (50%), Tigre; (30%), Saho and Afar (10%), Hedareb and Bilen (4.5%), Kunama, Nara and Rashaida (4%). An estimated 82% of Eritrea’s population lives in rural areas, subsisting through agriculture and raising livestock. Only about 20% of Eritreans are literate (source: www.cia.gov/cia/publications/factbook/geos/er.html).

Most of the nine ethnic groups speak Semitic or Cushitic languages. The Tigrinya and Tigre people make up four-fifths of the population. In general, most of the Christians live in the highlands, while Muslims and adherents of traditional beliefs live in lowland regions. Tigrinya and Arabic are the most frequently used languages for commercial and official transactions, but English is widely spoken and is the language used for secondary and university education (source: www.cia.gov/cia/publications/factbook/geos/er.html).

The National Assembly formally elected Isaias Afwerki, secretary general of the EPLF, president in June 1993. Isaias Afwerki remains the president of the country of what is described a nominal constitutional democracy. Democratic elections have been postponed indefinitely until the border dispute with Ethiopia is resolved.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-7
	October 2004

4.3.5

Mining Industry and Legislation

Eritrea currently has no operating metal mines.

During Italian colonial times, metal mining occurred at sites such as at Okreb and Augaro ceased once the British took control in 1941. Mining was conducted for gold at Augaro on a very limited basis in the 1950’s.

The Ethio-Nippon Company was mining the Debarwa (Cu, Pb, Zn) massive sulphide deposit in the early 1970's and the ore was sent directly to Japan for refining. Operating mines, such as Debarwa, were halted in 1974 as hostilities between the governing Ethiopian regime and Eritrean independence groups increased.

In 1995 the Eritrean government presented the Proclamation to Promote the Development of Mineral Resources (No. 68/1995) in association with the Regulation of Mining Operations (Legal Notice 19/1995; see Section 4.3.6). Additional regulations and proclamations have been presented regarding environmental protection, land use, water use and heritage (see Section 4.3.7).

4.3.6

Mineral Property Title

The State of Eritrea has provided several key documents relating to mineral property title and regulations.

Property titles are granted in Agreements with the State of Eritrea under the provisions of Proclamation No.68/1995 a Proclamation to Promote the Development of Mineral Resources.

Licences are granted and identified according to the level of exploration work completed on a property. Properties are granted under the following license types: Prospecting Licenses, Exploration Licenses or Mining Licenses. Properties can be obtained under one type of license and can be converted to the subsequent type if all obligations are met and the titleholder is not in breech of any provisions of the Proclamation and the appropriate application (with fees) are submitted.

The license entitles the licensee a 90% interest and the State of Eritrea holds the remaining 10% interest.

Properties are obligated to yearly expenditures depending on the size of the property and as described within the Agreement for the license. Annual License Rental fees for a property are based upon a levy of 200 Nacfa per km² with payment due on the anniversary of the license agreement for the property.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-8
	October 2004

Under the Regulation of Mining Operations (Legal Notice 19/1995) the holder of a Mining License shall pay the Eritrean government:

• Royalty for all minerals produced (see below).
  

• Income tax in accordance with the Proclamation No.69/1995.
  

• License renewal fee.
  

• Annual rental fees for license areas (as described above).
  

Additionally, the holder of a license and his contractors shall pay a 0.5% customs duty on all imports into Eritrea of equipment, machinery, vehicles and spare parts (excluding sedan style cars and their spare parts) necessary for mining operations.

The royalty to be paid by a licensee pursuant to Article 34 (1) of the proclamation shall be as follows:

a)

For precious minerals the royalty is 5%.

b)

For metallic and non-metallic minerals including construction minerals the royalty is 3.5%.

c)

For geothermal deposits and mineral water the royalty is 2%.

Notwithstanding this law, a lesser rate of royalty may be provided by agreement with the licensing authority, when it becomes necessary to encourage mining activities.

Taxation rates are described in the Proclamation No. 69/1995 Proclamation to Provide for Payment of Tax on Income from Mining Operations. A holder of a mining license shall pay income tax on the taxable income at a rate of 38%. Taxable income is to be computed on a historical accrual accounting basis by subtracting from gross income for the accounting year by taking into consideration all allowable revenue, expenditure, depreciation, re-investment deduction and permitted losses.

If any licensee transfers or assigns, wholly or partially, any interest in the license, the proceeds shall be taxable income to the extent that such consideration exceeds the amount of his un-recovered expenditure.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-9
	October 2004

Withholding taxes and personal income taxes of non-residents of Eritrea are identified within the proclamation. If the licensee contracts a company or person, who are not resident in Eritrea for services in Eritrea, the licensee will pay taxes on behalf of such a person. Taxes will be paid at the rate of 10% on the amount paid. For the purposes of this article in the proclamation, a person is temporarily present in Eritrea if he performs work in the country for more than 183 days in any accounting year. The compensation received by an expatriate employee of the licensee or his contractor shall pay an income tax at a flat rate of 20%.

The holder of a Mining License producing exportable minerals can open and operate a foreign currency account in Eritrea and retain abroad a portion of his earnings to be able to pay for importation of machinery, pay for services, for re-imbursement of loans and for compensation of employees and other activities that may contribute to enhancement of the mining operations.

The Government has reminded artisanal miners through Legal Notice No. 25/1995 Regulations for the Payment of Income Tax of a holder of an Artisanal Mining License to pay their income tax based on Article 22/1 of Proclamation No. 62/1994.

On September 2, 2004, Nevsun Resources Ltd. (NSU/TSX), Sanu Resources Ltd. (SNU/TSXV) and Sunridge Gold Corp (SGC/TSXV) each received a letter from the Minister of Energy and Mines for Eritrea instructing the companies to halt all mineral prospecting and exploration work and related activities in Eritrea until further notice. No reason was given for this instruction in the letter. Nevsun has temporarily suspended fieldwork in response to the stop work order.

AMEC is not aware of the circumstances or potential impacts of the instruction from the Ministry of Energy and Mines.

4.3.7

Environmental Regulations

Environment

In the absence of legislation to co-ordinate and manage the issues related to the environment, the Ministry of Land, Water and Environment has introduced The National Environmental Assessment Procedures and Guidelines (NEAPG) for undertaking environmental impact assessment for all development projects. The NEAPG provides mechanisms for ensuring an integrated approach to sustainable development.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-10
	October 2004

Land Use

Land Use regulations are described in the Land Proclamation, No.58/1994 which provides that all land is owned by the State and citizens have use right only. Under this Proclamation peasant farmers have the right to use land for a lifetime and if significant investment has been made on the land then priority is given for closer relatives to inherit the property and to continue farming the land. This proclamation has not yet been implemented, at least with respect to land distribution to peasant farmers. Legal Notice No.31/1997 was introduced to speed up the land law implementation process, which provided the legal basis for methods of land allocation and land administration. This Legal Notice mandates the Ministry of Land, Water and Environment, in collaboration with other ministries; to prepare land use and area development plans. The plans are still pending due to institutional and technical limitations.

Water Resources

The Ministry of Land, Water and Environment (Water Resources Department) has drafted a Water Law and efforts are being made to finalize and have it pass into legislation. The draft law deals with the institutional and regulatory issues, water use, water rights, and environmental and water quality. Currently water use is subject to the overlapping of water development interests of the Ministries of Agriculture, Public Works and local Government.

National Heritage

There is no integrated law that deals with National Heritage. The Cultural Assets Rehabilitation Project (CARP) has made studies on various aspects of National Heritage in Eritrea and has drafted a National Heritage law and efforts are being made to finalize and have it pass into legislation. The draft law deals with institutional and regulatory issues, heritage sites, preservation and rehabilitation.

The National Museum, which forms an integral part of the University of Asmara has the responsibility to educate the public; conduct research into critical issues that pertain to Eritrea’s past, its natural history, its social configurations, and its social and military history. The museum must also manage its diverse collections and is responsible for management of heritage sites (natural and cultural) and on-site museums, the dispensation of advice to owners of heritage objects and the enforcement of laws and regulations pertaining to heritage resources of all kinds.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 4-11
	October 2004

5.0

ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY

5.1

Accessibility

Asmara is the capital city of Eritrea and is serviced by regular international flights by Lufthansa Airlines out of Frankfurt; Regional Airways (subsidiary of British Airways) via Nairobi; and Eritrean Airlines servicing Amsterdam, Rome and Frankfurt.

Access to the Bisha Exploration License is by paved road from Asmara to Akurdat³, a distance by road of 181 km (Figure 4-1 and Table 5-1). From Akurdat access is via an all-weather compact dirt road to Adi Ibrahim (28 km) where the road degrades to a rough all-weather dirt road to the village of Mogorayb, which is the administration centre for the Dige Sub-zone. The Bisha permanent camp (Photo A-1 in Appendix A) is located 5 km south of the village of Mogorayb beside the Mogorayb River, and approximately 1.5 km north of the Bisha Exploration License boundary (Figure 5-1). The main work site at Bisha is located 4 km to the south of the camp along a dusty track across a flat alluvial plain (Photo A-2 in Appendix A).

Table 5-1
Distances by Road to the Bisha Exploration License

From	To	Distance (km)	Condition
Asmara	Akurdat	181	Paved, all weather road
Akurdat	Adi Ibrahim	28	Dirt, all weather road
Adi Ibrahim	Mogorayb	19	Dirt, rough, all weather road
Mogorayb	Bisha Camp	5	Dirt, all weather road
Asmara	Bisha Camp	233	4 hour drive
Bisha Camp	Main Gossan	4	Dirt, all weather road

The drive from Asmara to the Bisha camp is approximately 4 hours.

³ Also spelled “Agordet”.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 5-1
	October 2004

Figure 5-1
Property Access and Topography

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 5-2
	October 2004

5.2

Climate

The climate in the area is semi-arid with elevated temperatures year-round. During the hot season in April and May the average temperature is +42ºC, although temperatures may rise to +50ºC for short periods. The main rainy season is between June and September, and periodic flooding of the Mogorayb and Barka Rivers can result in spectacular flash floods. Occasional rain may also fall during April and May. Total rainfall is sparse with between 300 and 500 mm falling in the year.

The rainy season causes periodic, short-lived difficulty in travel off of the main highways, although exploration work is possible year round. During the period of exploration work by Nevsun the precipitation has only occasionally been sufficient to flood the local rivers (pers. comm. Ansell, 2004).

5.3

Local Resources and Infrastructure

There are few local resources in the Bisha area and the infrastructure is also limited.

The village of Mogorayb is the local administration centre for the Dige Sub-zone within the Gash-Barka District. The village has a small refugee re-settlement site and subsidiary military and commercial interests. The village contains a well-equipped, eight person health centre capable of taking care of small medical problems by nursing staff in preparation for referral of patients to larger, better equipped hospitals in Akurdat and Keren. Camp Mogorayb is a military training site located just outside the village boundaries. With the presence of the advanced exploration project at Bisha this camp has been re-activated as a security post from its previous care/maintenance basis.

Few basic goods are commercially available in the region, either in Mogorayb or Akurdat. The main centre for support of exploration and project development is from the capital city, Asmara.

The local population has no exploration or mining culture. Even workers from the larger populated centres would require training.

Water resources are very limited and any future mining operation would be reliant upon groundwater and the optimal use of recycled process water. The water table in the license area is very shallow during the wet season, being 1 to 2 m deep. During the dry season the water table can extend to 40 m in depth.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 5-3
	October 2004

Dirt roads and tracks cross the property but no paved roads exist south of Akurdat. A railroad bed crosses the property (Figure 5-1) but it is not continuous and the track was removed. The nearest telephone and electrical services are available in the town of Akurdat. Nevsun has diesel generators and satellite telephone service at the Bisha camp.

The principal port for importation of heavy equipment would be Massawa on the Red Sea coast, which is 348 km by road to the east (Figure 4-1), or approximately 7.5 hours driving time.

5.4

Physiography, Flora and Fauna

Physiographically, the area is mostly an alluvial plain at 560 masl (Figure 5-1) along the western foot of the Central Highlands. The Bisha, Wade and Neve peaks reach elevations of up to 1,226 masl above the alluvial plain at the southern boundary of the Bisha property.

The Bisha Exploration License is located on a flat to rolling desert-like plain that is typically desert with scattered vegetation and few trees. Steep hills and ridges rise above the plain (Photos A-1 to 9 in Appendix A). The soil regolith is made up primarily of 1.0 to 5.0 m of alluvial and eluvial material.

Abundant seasonal streams cross the area and flow northward from the Property into the Barka River (6 km north of the Bisha Exploration License boundary) and continue north and northeast into Sudan. The Bisha Exploration License is crosscut by the Mogorayb River, a tributary to the Barka River that flows northwards along the western side of the Property (Figure 5-1). A smaller seasonal tributary, the Fereketta River, flows north-northwest into the Mogorayb River. The Fereketta River crosses the Bisha Property and passes immediately west of the Bisha Gossan Zone (Photos A-5, 6 and 8 in Appendix A).

The area is covered by very limited and sparse vegetation and trees (see Photos A-2 to 6 in Appendix A). No specific details of fauna were available to AMEC, however ostrich, jackals, caracal, hyena and miscellaneous desert dwelling insects and reptiles have been observed in the region (pers. comm. Ansell, 2004).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 5-4
	October 2004

6.0

HISTORY

Nevsun has no record of any previous exploration or mining activities on the Property or surrounding areas prior to the 1996. In late 1996, Amanuel Woldu, an Eritrean geologist working for Ophir Ventures Inc. (Ophir), a private Canadian company, conducted prospecting in the Bisha area and collected samples from the gossanous outcrops (Table 6-1). Although this work resulted in the discovery of the surface exposure of the Bisha Deposit in the Bisha Gossan Zone, the actual deposit was not recognized until drilling commenced in 2002.

In late 1997, Ophir presented the property data to Nevsun and in early 1998 Bill Nielsen (V.P. Exploration) carried out a brief property examination. Nevsun optioned the property from Ophir. On June 3, 1998 Nevsun signed the Bisha Area Prospecting License Agreement with the State of Eritrea.

Table 6-1
General History of Bisha Property

Year	Company	Description
1996	Ophir Ventures	Prospecting, mapping and sampling
1998	Nevsun	Property examination and acquisition
June 3, 1998	Nevsun	Bisha Area Prospecting License Agreement signed
1998	Nevsun	Mapping (1:50,000), geochemical stream sediment sampling
June, 1999	Nevsun	Prospecting License converted to an Exploration License
1999	Nevsun	Geophysical surveys, mapping, geochemical sampling
1999	Phelps Dodge	Property examination.
1999 – 2002	State of Eritrea	Work suspended due to border war with Ethiopia – Force majeure
2002	Nevsun	Drilling of discovery outcrop area, mapping (1:1000)
2003	Nevsun	Phase I - Drilling, trenching, geophysics (airborne and ground), mapping, geochemical sampling, metallurgical testing, bulk density measurements
2003	Nevsun	Phase II - Drilling, geophysics, geochemical sampling, metallurgical testing, petrographic work, bulk density measurements
2004	Nevsun	Drilling (DDH and RC holes), geophysical surveys, mapping, geochemical sampling, petrographic work, bulk density measurements, geotechnical, environmental, metallurgical testing
September 2, 2004	State of Eritrea	Suspension of field work due to instructions by Ministry of Energy and Mines

In 1998, Nevsun completed reconnaissance scale geological mapping and a multi-element stream sediment sampling survey. The multi-element analyses defined anomalous base metal values in gossanous areas on the north side of the Bisha Area Prospecting License.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 6-1
	October 2004

In June 1999, the Prospecting License was converted to an Exploration License covering an area of 49 km².

In 1999, a grid was established over the gossan area of the Bisha Main Zone and geological mapping (1:5000), ground geophysical surveys (MaxMin, magnetometer) and limited “orientation” soil sampling was completed. Soil sampling showed the Bisha Gossan Zone to be highly anomalous in lead with significant values of copper, zinc and silver. Assays for gold were not completed for these samples.

Phelps Dodge Corp. completed a property examination in late 1999 and collected 10 grab samples (Table 6-2) of the gossan material. The samples returned anomalous gold values ranging up to 30.4 g/t Au. Samples B11 and B12 were collected from gossan outcrops 1.5 km northwest of the Bisha Gossan Zone, which correspond to the Northwest Zone.

Table 6-2
Phelps Dodge Corp. Grab Samples 1999

Sample #	Au	Ag	Cu	Pb	Zn	Mn	Co	Ni	As	Ba	Bi
B1	9.33	0.5	1147	465	754	1425	29	68	102	754	74
B2	1.01	0.5	985	499	609	2571	45	58	60	656	63
B3	0.75	1	1302	807	783	1534	77	52	1310	457	78
B4	3	0.5	599	1362	365	776	20	49	1214	361	82
B5	2.19	2	742	865	544	822	15	49	1916	367	84
B6	4.04	37	340	5631	338	762	16	71	2983	477	276
B7	0.39	4	1039	3462	449	1456	23	69	1949	565	82
B8	1.52	8	909	2275	391	606	17	64	1819	387	80
B9	30.4	18	2049	10135	551	1480	26	57	4755	374	101
B10	1.91	0.5	646	556	298	1044	45	62	207	525	71
B11	0.11	0.5	240	162	59	1276	101	87	806	1160	63
B12	0.08	1	503	191	62	921	13	59	868	780	66

Work was suspended between 1999 until 2002 due to the border war with Ethiopia. This was a force majeure event and therefore expenditure obligations and rental fees were also suspended during this time.

In October 2002, Nevsun completed diamond drilling program of 6 holes totalling 810.90 m at Bisha in order to test the geophysical and geochemical anomalies at the gossan outcrop area (Table 6-3). The drilling was sufficient to confirm the presence of a volcanogenic massive sulphide deposit overlain by a supergene copper-enriched zone and a gold-enriched gossan cap.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 6-2
	October 2004

Table 6-3
2002 Drill Program Summary of Significant Assay Intervals

Hole #	From	To	Interval (m)	Au g/t	Ag g/t	Cu %	Pb %	Zn %
B-2	29.00	66.00	37.00	0.02	0.65	0.93	0.00	0.00
B-3	4.35	14.50	10.15	1.96	1.72	0.07	0.06	0.04
B-3	19.0	28.96	9.96	10.24	44.80	0.07	1.95	0.04
B-3	134.95	172.00	37.05	0.99	24.94	0.97	0.04	1.92
Incl	134.95	155.00	20.05	1.46	40.4	1.52	0.06	3.09
B-4	48.77	56.39	7.62	5.44	88.53	0.10	0.92	0.03
B-4	56.39	101.20	44.81	0.87	27.13	3.92	0.11	0.34
B-5	37.50	45.72	8.22	8.53	693.35	0.06	9.65	0.01
B-5	45.72	57.00	11.28	16.52	475.32	3.62	8.28	0.02
Source: Chisholm et. al. (2003)

Two phases of diamond drilling were completed in 2003. The Phase I work was completed between February and June and consisted of diamond drilling 48 holes totalling 6,724.76 m, plus mapping, sampling, trenching, geophysics (airborne and ground), mapping, metallurgical testing, and bulk density measurements. The Phase II work was conducted between September and December and consisted of 93 core holes totalling 11,894.50 m. Additional work conducted during this program included geophysics, geochemical sampling, metallurgical testing, petrographic work, and bulk density measurements. A summary of the significant drill hole intersections of mineralization is provided in Appendix C (Table C-2).

Further diamond drilling (163 holes totalling 28,879.50 m), RC drilling (33 holes totalling 1,814.40 m) and core/RC combination holes (9 holes totalling 591.60 m) were completed between January and June 2004. Additional work completed during this program included geophysical surveys, mapping, geochemical sampling, petrographic work, bulk density measurements, geotechnical work, environmental baseline work, and metallurgical testing.

On September 2, 2004, Nevsun Resources Ltd. (NSU/TSX), Sanu Resources Ltd. (SNU/TSXV) and Sunridge Gold Corp (SGC/TSXV) each received a letter dated from the Minister of Energy and Mines for Eritrea instructing the companies to halt all mineral prospecting and exploration work and related activities in Eritrea until further notice. No reason was given for this instruction in the letter. Nevsun has temporarily suspended fieldwork in response to the stop work order.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 6-3
	October 2004

7.0

GEOLOGICAL SETTING

The regional geology of Eritrea and the adjacent countries of the Horn of Africa are not well documented and geological mapping within Eritrea has been limited due to the armed conflicts since the 1960’s. Recent country-scale mapping has been completed using LANDSAT imagery supplemented by limited field verification mapping. Portions of the regional geology in this report are summarized from a compilation by Chisholm et. al. (2003).

7.1

Regional Geology

Eritrea is underlain by the western or Nubian portion of the Arabian-Nubian Shield (Alemu, 2002; Figure 7-1). The exposure of this Precambrian greenstone belt is related to early stages of doming before opening of the Red Sea (Chisholm et. al., 2003) and is postulated to be the northern extension of the Mozambique Precambrian Belt (Berhe, 1990 in Chisholm et. al., 2003). The Red Sea is a post-Jurassic extensional feature.

Figure -1
Geology of the Arabian-Nubian Shield in the Red Sea Region

Source: http://www.utdallas.edu/~miller/SedGeoChem/snowball.html

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-1
	October 2004

The Arabian-Nubian Shield is composed of accreted Archean and Proterozoic rocks, which were reactivated during the Pan-African Orogeny in the Late Proterozoic-Early Paleozoic Era (1,000 to 500 Ma; Berhe, 1990 in Chisholm et. al., 2003). Granitoids intruded and metamorphosed older rock sequences thus resetting the geochronologic clocks to Pan-African dates and leaving a Pan-African structural overprint.

The age of the volcano-sedimentary rocks in the Arabian-Nubian Shield is not well known. In Eritrea, the units are considered to be approximately 850 Ma for the Tsaliet Group volcano-sedimentary rocks and >650 Ma for the overlying Tambien Group sedimentary rocks.

A geologic terrane map of the State of Eritrea in shown in Figure 7-2.

Figure 7-2
Geological Terrane map of Eritrea and Nevsun Exploration Licenses

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-2
	October 2004

Eritrea is divided into several north or northeast trending Proterozoic terranes, which are separated by major crustal sutures visible on Landsat satellite imagery (Berhe, 1990 in Chisholm et. al., 2003). The terranes are, from west to east: Barka Terrane, Hagar Terrane and Nacfa Terrane, Arag Terrane, and Danakil Terrane (see Figure 7-2).

Barka Terrane is comprised of metasediments that have been subjected to polyphase deformation. The terrane is bounded to the east by the Barka Suture, which follows the north-northeast trending course of the Barka River valley. The Hagar Terrane is east of the Barka Suture and is composed of ultramafics, olistostrome sediments within a volcano-sedimentary layered sequence (in Chisholm et. al., 2003). Berhe (1990) considered this to be a possible ophiolite sequence.

The Hagar Terrane was thrust into contact with the Nacfa Terrane. Nacfa Terrance is comprised of low-grade metamorphosed calc-alkaline volcanics and sediments. The Nacfa Terrane volcanics are considered by Drury and Berhe (1993) to be representative of a back-arc island arc to the Hagar Terrane. The Bisha Property is located within the Nacfa Terrane (Figure 7-2).

A compilation of studies on regional structures and mineral deposits by Chisholm et. al. (2003) is relevant to the mineral potential in the Nacfa Terrane as follows:

The Nacfa Terrane is likely a western extension of the relatively well-mapped Asir Composite Terrane of Saudi Arabia. The Asir Terrane is significant as it hosts numerous gold-base metal deposits in several north-south trending belts including the Wadi-Bidah and Wadi Schwas Mining Districts. Several of the Asir deposits are now coming to production and may in future be used as analogues to the known Eritrean deposits.

Chewaka and DeWit (1981) described the influence of plate tectonics on metallogenesis within Ethiopia and defined the Augaro and perhaps the Bisha area as part of a narrow, linear, north-south belt called the “Western Ophiolite Suture Zone” which extends north for many hundreds of kilometres from Western Ethiopia and continues north along the Barka River valley and northwards into Sudan and thence into the Red Sea. The zone is reported by them to be defined by aeromagnetic highs, LANDSAT lineaments and intermittent outcrops of gabbroic and ultramafic bodies.

A compilation by the Department of Mines of Eritrea (2003) has provided a more detailed summary of the distribution of lithologies within Eritrea (Figure 7-3). The Bisha Property is mapped as being covered by alluvium, and underlain by gabbro, and a volcano-sedimentary unit with very low-grade metamorphism.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-3
	October 2004

Figure 7-3
General Geology Map of Eritrea

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-4
	October 2004

7.2

Structural Interpretation of Western Eritrea

Major structural features in western Eritrea include terrane sutures and shear zones such as the north-south trending Barka River Fault or Suture (Figure 7-4). Preliminary structural interpretations of western Eritrea based on LANDSAT images (Drury and Charlton, 1990 in Chisholm et. al., 2003) have been used to support numerous other linear features. Chisholm et. al. (2003) compiled the interpretations with the unpublished 1:250,000 scale “Gash Area” geology map by the Department of Mines to produce an interpretation provided in Figure 7-4.

Several of the large-scale features identified in the compilation are also noted on the property scale maps. The key structural features relevant to the Bisha Deposit are the location: (1) adjacent to the Barka River Fault; and (2) on the northeast trending fold axis of a regional anticline. The deposit is also located to the northwest of a large gabbroic intrusion, which is a major feature on the property scale geology (see Section 7.0).

The Barka River Fault is a 200 km long north-south structure, which extends from southern Eritrea, through the Augaro (see below) and Bisha areas, northwards to the Sudan border.

The anticline in the Bisha area appears to plunge shallowly to the southwest near the Sudan border. The limbs of the anticline can be traced by marble units that act as a distinct marker horizon of purple-brown coloured ridges. The marble units have been mapped as the Fanco and Gogne Groups (Chisholm et. al., 2003).

7.3

Mineral Deposits of Eritrea

Mineral deposits have been identified in two areas of Eritrea, around the city of Asmara and also in the Gash-Barka District. Both of these areas are within the Nacfa Terrane.

A series of stratiform base metal occurrences and deposits were discovered in the Asmara Area within the Neoproterozoic age (854 Ma) Tsaliet Group rocks. The deposits are stratiform and occupy a series of three, subparallel 025º trends, which are, from west to east, Emba Derho⁴, Debarwa⁵, and Kodado⁶. The trends are separated laterally by a distance of 7 and 5 km, respectively from west to east.

⁴ Emba Derho trend hosts the Emba Derho deposit, Woki Duba (base metals + Au) occurrence, and Jerkeka gossan.

⁵ Debarwa Trend hosts the Adi Nefas deposit, Lamza Saharti gossan occurrence, Shiketi occurrence, Debarwa North deposit, Debarwa South deposit, and the Katina Occurrence.

⁶ Kodado Trend hosts the Kodado gossan and Adi Rassi deposit.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-5
	October 2004

Figure 7-4
Structural Interpretation of Western Eritrea

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-6
	October 2004

The deposits have been described as both Kuroko type deposits (Chewaka and DeWit, 1981) and as “bi-modal mafic type VMS” deposits (Sub-Sahara Resources web site; www.subsahara.com.au). The Asmara deposits are hosted by a variety of sedimentary and volcanic rock types. Barite is a common constituent of the deposits, which further supports the classification as volcanogenic massive sulphide deposits.

The Debarwa massive sulphide deposit was partially mined by the Ethio-Nippon Company in the early 1970's. Both Debarwa and Adi Nefas deposits are currently being explored by Sub-Sahara Resources (Sub-Sahara Resources web site; www.subsahara.com.au).

The deposits typically have a hematite/goethite gossan on surface ranging from a few metres up to 50 m vertical thickness. The gossans extend along the strike of the deposits, and at Debarwa the gossans outcrop over a 1,500 m strike length. The Debarwa deposit gossan (approximately 50 m thick) is underlain by a 30 m thick chalcocite blanket, which is in turn underlain by primary chalcopyrite mineralization. The mineralization dips at roughly 55º to the west and consists of a single main zone underlain by a thin footwall zone.

Mineralization consists of pyrite, chalcopyrite ± galena in the primary zone and the supergene zone consists of chalcocite, covellite, digenite, bornite and tennantite (Sub-Sahara web site; www.subsahara.com.au). Recent exploration by Sub-Sahara Resources has encountered zinc-rich mineralization at surface, 200 m to the north of the Debarwa deposit.

Mineral resource estimates were prepared for several of the deposits (Table 7-1) by BRGM-La Source, Phelps Dodge Corp. or the African Mineral Resources Development Centre.

Table 7-1
Asmara Area Base Metal Prospects and Deposits

Trend	Deposit	Tonnage (Mt)	Cu %	Pb %	Zn %	Au g/t	Ag g/t
Emba Derho	Emba Derho	2.5 resource7	0.39	-	2.36	0.10	4.5
Debarwa	Adi Nefas Zinc	1.43 resource8	0.95	est. 0.8	9.30	3.28	129.0
Debarwa	Adi Nefas Doop	2.929	-	-	-	3.10	-
Debarwa	Debarwa	1.65 mineable reserve10	5.1		-	1.40	-
Debarwa	Debarwa South	1.29 inferred resource11	5.1(?)	-	-	1.40(?)	-
Kodado	Adi Rassi	4.5 resource12	1.0	-	-	0.70	-

⁷ Phelps Dodge Corp.

⁸ African Mineral Resources Development Centre quoted in Sub-Sahara web site.

⁹ BRGM – La Source (1996-7) quoted in Sub-Sahara web site.

¹⁰  Phelps Dodge Corp. in Sub-Sahara web site.

¹¹  Phelps Dodge Corp. in Sub-Sahara web site.

¹²  Source not specified in Chisholm, 2003.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-7
	October 2004

AMEC has not confirmed whether the estimates listed in Table 7-1 meet the criteria of NI 43-101 and therefore are reported for reference purposes only. AMEC cannot confirm the reliability or relevance of the estimates therefore AMEC makes no warranty as to their validity or accuracy.

The series of gold and base metal deposits noted in the Asmara Area continues southwest 160 km into northern Ethiopia to the Tsehafi Emba deposit. Also, the base metal discoveries near to Asmara are located in close proximity to a large number of small gold occurrences in the Harassing gold camp including a number of historical gold producers such as Medicine, Hara Hot, Sciumagalle and Adi Nefas Doop. Some of these prospects have been the source of artisanal gold production.

The current focus of exploration in Eritrea is the deposits and prospects in the Augaro-Bisha Area. The VMS base metal and gold-rich mineralization at Bisha Main Zone and the Northwest Zone has generated interest for exploration companies. Recent exploration license applications have focused on the western Nacfa Terrane along t he Barka River Fault.

Few details are available for gold production during the Italian colonial times from the Augaro mine. The property is located to the south of Bisha and is also held as an Exploration License by Nevsun.

A number of other areas in Eritrea have potential for gold-base metal deposits. These include the Beddaho, Raba and Semait areas of northern Eritrea and the Shambiko area on the Ethiopian border east of Augaro. Eritrea is under-explored and has potential for further mineral deposits within the Nacfa and other terranes.

In addition to the mineral deposits of Eritrea, the deposits in northwestern Ethiopia, southeast Sudan and western Saudi Arabia should also be discussed because of the mineral potential of the Nubian portion of the Nubian-Arabian Shield, which extends into these countries (see Table 7-2).

Several base metal and gold deposits have been discovered in the Proterozoic greenstone belts of Sudan. The principal mineral deposits include the Hofrat en Nahas, Hassai and Oderuk, which are the main sources of gold production in Sudan (Table 7-2). The deposits are all stratabound VMS-style mineralization. Recently the Sudan geological survey (GRAS) estimated that the primary VMS deposits of Hassai, Hadal Auatib, Oderuk have a combined total of approximately 62 Mt.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-8
	October 2004

Table 7-2
Mineral Deposits in Eritrea, Sudan, Ethiopia and Western Saudi Arabia

Country		District	Deposits	Metals	Comments
Eritrea		Asmara	Emba Derho, Debarwa North, Debarwa South, Adi Nefas, Adi Rassi	Base metals, Au	VMS
		Gash-Barka	Augaro	Au	Shear zone hosted gold deposit
		Gash-Barka	Bisha	Au, base metals	VMS
Ethiopia		Asmara (160 km SW)	Tsehafi Emba	Cu, Au	VMS
Sudan			Hofrat en Nahas	Cu	Sedimentary copper
		Abu Samar			Stratabound VMS
		Ariab	Hassai, Oderuk, Hadal Auatib, Adaiamet	Au, Cu, Zn	Stratabound VMS with gossan zone, reserve of 3.57 Mt at 9.1 g/t Au in 1999, combined resource of 62 Mt, open pit mining, approx.350,000 oz produced per year, over 2 Moz produced to date
Saudi Arabia		Jabal Sayid	Jabal Sayid	Au, Cu	Weathered base metal deposits, VMS, Kuroko, Resource of 20 Mt @ 2.68% Cu within larger resource of 80 Mt @ 1.5% Cu
		Jabal Sayid	Mahd Ad Dahab	Au, Ag, Cu	Epithermal related to VMS, Mined since 961 BC, total deposit estimated to be 3.2 Moz Au, trackless UG and small open pit
		Wadi Schwas	Al Hajar	Au, Cu	Stratiform, Weathered base metal deposits, Similar terrane to Nacfa, reserves of 3.5 Mt @ 3.28 g/t Au and 38 g/t Ag, gossan+supergene+primary, annual production of 55K oz Au and 235K oz Ag from OP
		Wadi Bidah			VMS with surface gossan
	Source: Chisholm et. al. (1993)

AMEC has not confirmed whether the estimates listed in Table 7-2 meet the criteria of NI 43-101 and therefore are reported for reference purposes only. AMEC cannot confirm the reliability or relevance of the estimates therefore AMEC makes no warranty as to their validity or accuracy.

Numerous gold and base metal deposits have been discovered in Proterozoic-aged greenstone terranes in Saudi Arabia (Table 7-2). The Jabal Sayid and Al Hajar mines are currently in production and consist of weathered base metal deposits having a very high gold grade. These mines are located in western Saudi Arabia in Proterozoic arc terrains, which may correlate to the Hagar and Nacfa terranes of western Eritrea. The Barka River Fault in Eritrea is postulated (Abdelsalam et. al., 2003) to continue north into the Asir terrane of Saudi Arabia as the Bidah shear within the Wadi Bidah Mining District. The Wadi Bidah District is the location of over 16 different base metal massive sulphide deposits, six of which have significant gold content (Chisholm et. al., 2003). All of the VMS deposits are marked by significant surface gossans.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-9
	October 2004

The Jabal Sayid District includes two major deposits that are currently in production: the Jabal Sayid VMS deposit and the Mahd Ad Dahab epithermal deposit (see Table 7-2). Both are low cost producers of gold, silver and base metals.

The Mahd Ad Dahab deposit has ancient surface workings that were mined for over two thousand years, starting in 961 BC and it has been estimated that 1 million oz of gold and silver was extracted during this time. Modern production started in 1988 and ore mined to 2001 was reported (Ma’aden web site: www.maaden.com.sa) to be 2.731 Mt with a gold grade of 21.89 g/t along with unpublished silver and copper. The total size of the deposit has been estimated to be 100 t of gold (3.2 million oz).

The Al Hajar deposit is located within the Wadi Schwas district and may be a northern extension of the Nacfa Terrane greenstone terrane in Eritrea. Published reserves (Ma'aden web-site: www.maaden.com.sa) are 3.5 Mt grading 3.28 g/t Au and 38 g/t Ag. Supergene reserves have been identified within a gossan, which overlies a primary massive sulphide deposit. The Al Hajar Mine produces 55,000 oz Au and 235,000 oz Ag per year from an open pit operation with ore treated by heap leach cyanidation.

7.4

Property Geology

The Bisha Property geology was recently geologically mapped and compiled (including work by Daouda Ouattara, Robin Chisholm, Amanuel Woldu, Fiona Childe, and the late Tim Nutt) into a comprehensive map (Figure 7-6) with a corresponding detailed interpretation and 1:5,000 scale mapping over the Main Gossan Area (Figure 7-7; Greig, 2004). The following geology summary is largely derived from the latter work.

The Bisha Property is underlain by low-grade metamorphosed (upper greenschist to lower amphibolite facies) volcanics and sedimentary units on the western margin of the Nacfa Terrane (Figure 7 -6). The precious metal-rich massive sulphide deposits at Bisha are hosted by a tightly and complexly folded, intensely foliated bimodal sequence of generally weakly stratified, predominantly tuffaceous metavolcanic rocks (Greig, 2004). The Bisha Gabbroic Complex has intruded the south and central portion of the Property. Fold axis and marker horizons provide clear trends and enable development of a general stratigraphic and a structural model.

AMEC reviewed the surface geology, logging and geologic model used for resource modelling and considered these to be a reasonable representation of the stratigraphy and structure on the Bisha Property and of the Bisha massive sulphide deposit morphology.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-10
	October 2004

7.4.1

Stratigraphy

The sedimentary rocks consist primarily of greywacke, siltstone, shale, marble, and feldspathic arenites with less common conglomerate, magnetic ironstone, quartzite and massive sulphides lenses. The volcanic sequence includes fine-grained pyroclastic rocks of mafic to intermediate composition and pillowed mafic flows, felsic ash and lapilli tuffs.

Figure 7-5
Stratigraphic Section

The stratigraphic section in Figure 7-5 (Barrie, 2004) corresponds with the following summary of the Property stratigraphy provided by Greig (2004) for the map units¹³ as presented in Figures 7-6 and 7-7.

In general, stratified rocks at Bisha can be divided into two parts: an upper, predominantly felsic volcanic part that is capped by sedimentary rocks; and a lower volcanic part that is clearly bimodal, at least in the south and east. This lower bimodal volcanic part appears to be capped by the mineralized stratiform mineralized horizons at Bisha Main and Bisha South, while the mineralization at the Northwest Zone may be at a somewhat higher stratigraphic level.

The upper sequence of felsic metavolcanic rocks which hosts the deposits is best exposed to the north and northwest of the mineralized horizons at Bisha. It consists predominantly of tuffaceous felsic deposits, although rare mafic rocks, principally fine tuff (ash and fine lapilli tuff), do occur. Overlying these felsic tuffaceous metavolcanic rocks, and in part interbedded with them, are a sequence of fine-grained clastic rocks. These underlie the plains to the immediate south-southwest of the mineralized zones, as well as the more

¹³ The corresponding Nevsun logging codes were added by AMEC. References to the figures and photos provided in the original report have been removed.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-11
	October 2004

extensive low-lying areas farther to the west. They are generally very poorly exposed, but their presence in those areas, as well as to the north, has been inferred from the airborne geophysics. The lower part of the stratigraphy, occurring in the footwall of the mineralized horizons, generally underlies the area south and east of the mineralized zones, closer to the northwestern contact of the Bisha gabbroic complex. As mentioned, this part of the stratigraphy is bimodal, consisting of interbedded felsic (dacite, rhyodacite and rhyolite) and mafic (basalt, local basaltic andesite) metavolcanic rocks. In contrast to the areas to the north and west that are underlain principally by felsic rocks, there is a greater component of “proximal” deposits, such as flows and coarse tuffaceous rocks, in this area and of both felsic and mafic compositions.

In the sections below, the map units shown in the Figure 7-6 are described in general stratigraphic order, from youngest to oldest. As mentioned above, the five “undivided” map units mainly represent areas of scant map control–either there is sparse outcrop or little traverse coverage.

Map Units–Stratigraphic Rocks

Map unit “s” undivided metasedimentary and subordinate(?) fine-grained tuffaceous metavolcanic rocks. (Nevsun logging codes: SDST, CONG, QTBX, QUAR) As mentioned above, rocks assigned to this unit underlie the plains to the immediate south-southwest of the mineralized zones, as well as the more extensive low-lying areas farther to the west (Figure 7-6). They are generally exposed only in parts of a few seasonal drainages, but their presence in the surrounding areas, as well as to the north and in the area southwest of the Northwest Zone, has been inferred from airborne and local ground geophysics (generally good conductivity, low magnetic relief, subdued gravity signature). About 1 km south-southwest of the Bisha massive sulphide horizons, the rocks are predominantly olive green siltstone and fine-grained sandstone, with local medium-grained and rare coarse-grained green pebbly sandstone. The latter rocks in one place dip gently south, and scours at the base of a normally graded pebbly sandstone bed suggest that they are right-way-up. Pebbles in the sandstone appear to consist of felsic volcanic rocks. Farther to the south-southwest, well-foliated grey-green or grey siltstone and silty mudstone are more common, and locally, the rocks taken on a purplish or rusty weathering colour. In the very fine-grained rocks, bedding and foliation are commonly at a very low angle to one another. Toward the lower contact of the metasedimentary map unit, thin buff and locally rusty weathering ash and dust tuff layers may occur, and the contact with the underlying fine-grained felsic tuffaceous metavolcanic rocks is placed where the tuff beds become, in general, more abundant than the metasedimentary rocks.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-12
	October 2004

Map unit “vs” interbedded fine-grained felsic tuffaceous metavolcanic rocks and subordinate metasedimentary rocks. (Nevsun logging codes: SDST, CONG, QUAR) As with the predominantly metasedimentary rocks described above, rocks of this unit outcrop mainly to the south-southwest of the mineralized horizons at Bisha. They consist of similar fine-grained lithologies, but in different relative proportions, with felsic fine-grained tuffaceous rocks (dust and ash tuff) predominating over metasedimentary rocks; rare coarser-grained tuffaceous rocks also occur. The tuff is typically white to buff to rusty weathering and may or may not contain fine- to medium-grained quartz eyes.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-13
	October 2004

Figure 7-6
Property-Scale Geology Map

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-14
	October 2004

Figure 7-7
Deposit-Scale Geology Map

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-15
	October 2004

Map unit “g” “heavy” gossan (weathered massive sulphides). (Nevsun logging codes: SAPR, HEBX, FERU, FERC, HALF, SAND) At Bisha Main, and locally at Bisha South and the Northwest Zone, dark brown to black, very dense silica and iron-rich gossan boulders commonly occur at surface (Figure 7-6 and 7-7). Their density distinguishes them from gossanous rocks elsewhere on the property, and diamond drilling has shown that the boulders are the surface expression of deeply-weathered (typically several tens of metres) massive sulphide horizons. The gossans are locally traceable for well over 100 m, and at Bisha Main, they, together with a closely associated and extremely siliceous lithology (rhyolite(?); see below), outline the nose of the “Bisha syncline,” the major moderately south-plunging syncline which hosts the mineralization at Bisha Main and Bisha South (Figure 7-6).

Map unit “r” rhyolite, rhyodacite, and related rocks. (Nevsun logging codes: FELD) The mineralized horizons at Bisha Main, Bisha South, and the Northwest Zone are closely associated with rhyolite, rhyodacite, and related felsic metavolcanic flows, flow-breccias, and block tuff, although in general, the felsic rocks at Bisha appear to be finer-grained and of a less “proximal” nature. The high-silica flows and coarse fragmental deposits are typically resistant and are commonly marked by the presence of flow-layering or flow foliation as well as by the presence of abundant sulphides, typically pyrite. Weathering of these silica- and sulphide-rich rocks typically yields a hackly surface. The rhyolitic rocks in the vicinity of the massive sulphide mineralization appear to occur at more than one stratigraphic horizon (probably several, in fact), although, in general, the rhyolitic horizons appear confined to a stratigraphic interval within dacitic to rhyodacitic ash and lapilli tuff that is probably little more than a hundred or so metres thick at its thickest. More commonly, this stratigraphic interval is much thinner, and it is commonly identified only by the presence of local metre-scale blocks and layers of distinctive resistant, rusty-weathering grey rhyolitic rocks that contain abundant disseminated pyrite. The possibility that some of these pyritic rhyolitic rocks represent dykes or sills cannot be completely discounted, but the presence of rare interbeds of fine-grained well-bedded clastic and/or fine tuffaceous rocks, as well as the local presence of what are clearly fragmental rhyolitic rocks suggests that most are stratified as opposed to intrusive. The pyritic rhyolitic rocks are clearly discontinuous along strike, and as commonly as not, they are absent. For example, at the northern end of the Northwest Zone, resistant rhyolitic flows which comprise up to ten or more metres in stratigraphic thickness are essentially absent not much farther in either direction along strike than their stratigraphic thickness–this suggests that abrupt facies changes are the common with these rocks. Elsewhere, however, such as a short distance away along the eastern limb of the Northwest Zone syncline, the rhyolitic rocks form relatively continuous markers. There, distinctive, relatively resistant flow foliated rhyolite layers that are commonly less than a metre thick can be traced along strike more or less continuously for a little over a kilometre.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-16
	October 2004

Map unit “if” silicate-magnetite iron formation. (Nevsun logging codes: FERU, FERC, HALF). Strongly magnetic, dense and relatively resistant silicate-magnetite iron formation, possibly of exhalative origin, weathers a distinctive black to very dark brown colour, and typically has a mottled black, brown, and white colour on fresh surfaces (Figure 7-6). The iron formation occurs in a number of places in and around the Bisha property, including: 1) two localities to the east of the Northwest Zone syncline, near to and just north of the property boundary; 2) an area not far to the southwest of the Northwest Zone; 3) a locality in a drainage near the footings for a collapsed trestle on the abandoned railway line approximately 1.5 km southwest of Tabakin Range; 4) a locality approximately 500 m east of the central part of Tabakin Range; and 5) a locality approximately 1.5 km southeast of The Tabakin Range. The iron formation is thickest and most extensive at the latter locality, where it reaches a thickness of several metres and has a strike extent of approximately 350 m.

Although the origin of the silicate iron formation and its relationship to massive sulphide mineralization remain to established, the fact that the iron formation appears to be interlayered with metavolcanic rocks suggests that it is sedimentary in origin. In addition, the local presence within adjacent tuffaceous rocks of what appear to be irregular silica-magnetite “concretions”, around which the regional foliation is wrapped, indicates that the silica-magnetite rocks pre-dated the regional deformational event, and supports the interpretation that the these rocks were deposited contemporaneously with tuff and/or were formed shortly thereafter, via replacement processes.

Map unit “qt” quartz eye-bearing felsic fine tuffaceous metavolcanic rocks (mainly ash and fine lapilli tuff). (Nevsun logging codes: FELD, QFPD?, INTT?) Quartz eye-bearing felsic fine tuffaceous metavolcanic rocks, mainly ash and fine lapilli tuff, are only shown on Figure 7-6 in the Tabakin Range, which is one of the few areas on the property where the amount of outcrop and the level of detail in geologic mapping has been sufficient to distinguish individual lithologic units. Correlative rocks are common elsewhere on the property, however, and quartz-bearing fine tuffaceous rocks make up significant proportions, if not the bulk of, a number of other metavolcanic units shown in Figure 7-6, including units ufv (undivided felsic metavolcanic rocks), uv (undivided metavolcanic rocks), mft (interbedded mafic and felsic tuffaceous metavolcanic rocks), and afv (altered hematitic, sericitic felsic metavolcanic rocks).

The fine tuffaceous rocks are orange-brown weathering and buff, white, or pale grey on fresh surfaces. They are commonly well foliated, although typically not as well foliated as their quartz-poor counterparts (see below). The fine- and fine- to medium-grained quartz eyes which characterize these rocks in the field vary considerably in their abundance, from as little as a few percent to as much as 10 percent or more.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-17
	October 2004

Map unit “ufv” undivided felsic metavolcanic rocks. (Nevsun logging codes: FELD, FPDK) Much of the immediate host stratigraphy to the massive sulphides at Bisha Main, Bisha South, and the Northwest Zone are shown on Figure 7-6 as undivided felsic metavolcanic rocks. The felsic rocks consist mainly of fine-grained rhyodacitic or dacitic tuffaceous rocks (predominantly fine lapilli and ash tuff), but medium to coarse felsic lapilli and block tuffs are also locally common. However, because of the strong overprint of foliation in these rocks, and because of the lack of compositional contrast between fragments and matrix, it may be difficult to distinguish fragments from matrix in these rocks, and it is therefore possible that within the felsic metavolcanic map units the abundance of coarse fragmental rocks, or even flows, may be underestimated.

The most common lithology in the area immediate to the Bisha Property massive sulphide horizons, and particularly between Bisha Main and the Northwest Zone, is a moderately rusty brown weathering, pale to medium green ash to fine lapilli tuff that is characteristically well foliated. Primary stratification in these rocks is typically very difficult to identify, in part because of the overprint of foliation, but also because of the massive nature of what are most likely thick, poorly stratified tuffaceous deposits. Locally, however, dust to fine ash tuff beds display well-developed centimetre- to decimetre-scale bedding, although other fine tuffaceous rocks, such as the white- to buff-coloured ash and local fine lapilli tuff common in the southern part of the Tabakin Range, appears, in most places, to be at best only weakly stratified. The coarser tuffaceous deposits common to map units at Bisha that contain felsic rocks include coarse lapilli and local block tuff, which are particularly notable on the east side of the ranges east of the Northwest Zone. They also include fine to medium lapilli tuff, which is locally interlayered with finer grained tuffaceous rocks in various places on the property. Locally, such as in drill holes in the immediate vicinity of the massive sulphide horizons, the contrast between fragments and matrix in felsic lapilli tuff is highlighted by relatively intense chloritization of the fine-grained matrix, which was most likely originally rich in volcanic glass (i.e., hyaloclastite).

Map unit “ab” coarsely amygdaloidal metabasalt flows and associated meta- fragmental rocks. (Nevsun logging codes: MAFF) In the central and southern part of Tabakin Range, relatively massive amygdaloidal basaltic rocks form a distinctive map unit. The basalt is medium to dark green and commonly contains very coarse amygdales of up to 5 centimetres in long dimension. Amygdales are typically filled with quartz and/or calcite, and, at least locally, magnetite. Similar amygdaloidal basaltic rocks are also found interbedded with tuffaceous felsic metavolcanic rocks immediately east of the Tabakin Range within map unit “mft”, but have not been subdivided because of lack of map control.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-18
	October 2004

Map unit “umt” undivided mafic tuffaceous metavolcanic rocks. (Nevsun logging codes: MAFU, MAFT, INTT) As with the amygdaloidal basalts described above, these rocks have been broken out only in the Tabakin Range, where there is sufficient continuity of exposure. The fragmental rocks are dark green, basaltic in composition, and range from medium to coarse lapilli tuff. Although they appear on to be closely associated with the amygdaloidal basalt, they typically do not include fragments of it, and therefore appear to have emanated from a separate volcanic centre or during a separate eruptive event.

Map unit “mft” interbedded mafic and felsic tuffaceous metavolcanic rocks. (Nevsun logging codes: MAFT, FELD, INTT) Map unit “mft,” which mainly underlies the plains to the south and east of the Tabakin Range, consists of interbedded mafic and felsic tuffaceous metavolcanic rocks that appear, in large part, to occur in a similar stratigraphic setting to the rocks in the range itself. Individual mafic and felsic components of this unit have not been subdivided because of the limited exposures in this area–outcrops are limited to the few dry stream drainages which have cut deeply enough into the alluvial cover to expose bedrock. Mafic and felsic tuff appear to occur in more or less equal abundance in the unit, and in a few places, metre-scale felsic or mafic dykes and sills have intruded the stratified rocks. A common and very distinctive lithologic component of this map unit are highly altered felsic (and subordinate mafic(?)) metavolcanic rocks that commonly contain pale-weathering medium- to coarse-grained aluminosilicate(?) porphyroblasts (?). The porphyroblastic rocks were previously mapped as porphyritic andesite, but the porphyroblasts clearly overgrow tuffaceous fragments and the matrix is commonly rich in hematite, limonite, and sericite. The generally medium to dark rusty brown weathering rocks are relatively dense, and in the field, their colour and density yield the general impression that they are of mafic composition. Their trace and rare earth element geochemical signature, however, suggests that most, if not all, of these rocks are felsic, and it suggests further that pervasive Fe carbonate alteration and/or sulphidization (with the sulphides now weathered to Fe oxides) may have resulted in an increase in density.

Map unit “mqer” massive quartz eye-bearing metarhyolite and metarhyodacite flows and associated meta-fragmental rocks. (Nevsun logging codes: QFPD, QPDK, FBPD) Massive quartz eye-bearing metarhyolite and metarhyodacite flows and associated meta-fragmental rocks outcrop in the south and southeast parts of the Tabakin Range. They are generally foliated, but are commonly more blocky weathering than most other felsic metavolcanic rocks on the Bisha property. Fracture and joint surfaces in these rocks are typically coated with hematite and subordinate limonite, and hematite and limonite boxwork (after pyrite) is a common feature, although they are typically pale grey to white on fresh surfaces.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-19
	October 2004

Map unit “umv” undivided mafic metavolcanic rocks. (Nevsun logging codes: MAFF) Rocks of map unit “umv,” which represents undivided predominantly mafic metavolcanic rocks, underlie the area south and east of the mineralized zones at Bisha, in an area with little traverse coverage and/or outcrop. Lithologies in the areas traversed are similar to those in map units “umt” and, to a lesser extent “ab.”

Map unit “m” marble. (Nevsun logging codes: none) Marble outcrops in one place in the eastern part of the area mapped (Figure 7-6), near the contact with the Bisha gabbroic complex contact and approximately 3.4 km southeast of Bisha South. The marble occurs in a body approximately 200 m long and up to 5 or more metres thick and several other metre-scale carbonate bodies nearby are interbedded with well-foliated metabasalt. The carbonate is generally pale grey, and buff weathering, and locally contains centimetre- to decimetre-scale knots of calc-silicate skarn. Local intrafolial folds were also common within the body.

Map unit “uv” undivided metavolcanic rocks. (Nevsun logging codes: INTT, all other possible volcanic rock codes) Map unit “uv,” undivided metavolcanic rocks, represents areas that, on the basis of their airborne geophysical response, are most probably underlain by metavolcanic rocks, but in which there is either very little outcrop or little to no traverse coverage.

Adjacent portions of the Nacfa terrane have been covered by flood basalts dated at 30 Ma (Upper Paleogene) but these are not evident within the Bisha Property area.

A thin veneer of recent alluvial outwash sediments and talus covers more than half of the property. The outcrops and gossans occur as rocky ridges and isolated “islands” within the alluvial plain.

7.4.2

Intrusives

The region has been intruded by diorite and gabbro including the Bisha Gabbroic Complex; a very large gabbroic intrusive that forms large hills within and south of the current Bisha concession (Figure 7-6). Swarms of narrow, syn-tectonic felsic dykes oriented parallel to bedding were mapped cutting all major units. The dykes are felsic to intermediate in composition. Later post-tectonic intrusives include quartz-feldspar porphyry, granodiorite and quartz diorite.

The following descriptions¹⁴ of the intrusive map units are from Greig (2004) and correspond to the maps in Figure 7-6 and 7-7.

¹⁴ The corresponding Nevsun logging codes were added by AMEC. References to the figures and photos provided in the original report have been removed.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-20
	October 2004

A number of intrusions, both pre- and post-tectonic, are present within the area shown in Figure 7-6. Previous mapping on the property showed even more intrusive rocks, particularly in the southern part of the Tabakin Range, where relatively massive tuffaceous rocks were previously interpreted to be felsic intrusive rocks. Although this interpretation is now in disfavour, the possibility remains that some of these rocks are indeed intrusive, and further work is needed in that area to further our understanding. Intrusive rocks still underlie a significant proportion of the area mapped, particularly in the east and southeast, where the Bisha gabbroic complex and several unnamed intermediate to felsic plutonic bodies occur. They have not, however, been the focus for exploration, and therefore remain essentially unmapped–their presence has been inferred largely from the airborne geophysical data.

Map Units–intrusive rocks

Map unit “fd” felsic dykes. (Nevsun logging codes: FELD, FPDK, QFPD, QPDK, FBDK) Felsic dykes are common in the area approximately two kilometres east of Bisha Main, and outcrop locally in the Tabakin Range (Figure 7-6). In the latter area, they are non-foliated and granitic in composition.

Map unit “ifp”intermediate to felsic plutonic rocks. (Nevsun logging codes: INTD) Intermediate to felsic plutonic rocks shown in Figure 7-6 include one pluton approximately two kilometres east of Bisha Main that may be genetically associated with nearby felsic dykes, and a more extensive body to the south, along the contact of the Bisha gabbroic complex. The latter body clearly intrudes adjacent mafic metavolcanic rocks and is itself at least locally foliated, but its relationship to the Bisha gabbroic complex has not yet been established. A granitoid dyke in the Adalawat Range not far east of the Northwest Zone, intrudes foliated felsic metavolcanic rocks but is itself clearly nonfoliated.

Map unit “md”mafic dykes. (Nevsun logging codes: MAFD) Although smaller-scale mafic dykes occur elsewhere on the property, the only map-scale body is the massive fine-grained basalt dyke shown on Figure 7-6 near the south end of the Tabakin Range.

Map unit “bgc” Bisha gabbroic complex. (Nevsun logging codes: none) As mentioned above, rocks of the Bisha gabbroic complex underlie a significant portion of the area mapped, particularly in the east and southeast. Where examined, the rocks of the Bisha complex appeared to consist mainly of variably foliated amphibolitized(?) metagabbro to leucogabbro and/or diorite. In one locality, marble and mafic metavolcanic rocks adjacent to the contact possessed a relatively well-developed foliation relative to rocks farther from the contact, and the mafic rocks were largely recrystallized to fine- to medium-grained amphibolite, suggesting the possibility that the Bisha complex was emplaced at depth into

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-21
	October 2004

its volcanic host rocks. The presence of similar rocks to the Bisha gabbroic complex on the north and northeast (Figure 7-6) has been inferred from airborne magnetic data.

7.4.3

Structure

Sedimentary and volcanic rocks have a regional strike of north to northeast with local variation due to folding. Sedimentary and volcanic “tops-up” indicators and metal zoning with the massive sulphide deposit support the stratigraphy as being oriented “right-way up” (west-facing).

The stratigraphy on the Bisha Property is folded and comprises of a series of folds with axes plunging shallowly to the south and southwest. Although on the property-scale there are two clear fold axes (Figure 7-6), on the deposit scale the folding is much more complex (Figure 7-7). The folds are tight to isoclinal, nearly symmetrical with axial planes overturned slightly to the east. Additional smaller and/or parasitic folds are noted in several areas.

Folding has caused thickening of the massive sulphide intervals in the fold noses (and attenuation on some limbs) due to mass flow from the limbs.

The structural summary based on the property-scale and deposit-scale mapping by Greig (2004) is as follows:

The rocks at Bisha are characterized in hand specimen and at the outcrop-scale by the presence of a strong flattening foliation and a well-developed elongation lineation that appear to have been developed as an axial planar foliation to a series of variably plunging, tightly appressed folds. The folds and foliation essentially mirror the orientation of the contact of the Bisha gabbroic complex, suggesting that it may have acted as a more competent “indenter” during a phase of regional contractional deformation.

Folds

Although much less conspicuous on the outcrop scale than the well developed foliation common to the rocks at Bisha (see below), folds are perhaps the single most significant structural element on the property, and the regional foliation is developed, more or less, as an axial planar fabric to the map-scale folds. Fold geometries, however, may be complex. They are typically appressed, with narrow hinge regions and long limbs, and generally upright to slightly overturned. Axial trends are generally to the north-northeast or north, although in the northeast part of the area mapped, folds appear to trend to the north-northwest. It should be noted that this gradual change in orientation of the fold axial surfaces, on both an individual (e.g., the Northwest Zone syncline) and collective basis, mirrors the property-scale changes in trends of foliation, and in a general sense, it also

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-22
	October 2004

appears to mirror the trend of the contact of the Bisha gabbroic complex. Fold axes are variably plunging, and plunge reversals appear to be common; although marker units for outlining the resulting hinge line culminations are scarce, several domes and basins that reflect such culminations are clearly apparent. This is perhaps best displayed at the Northwest Zone, where a basin and the doubly-plunging Northwest Zone syncline are outlined by resistant rhyolitic rocks. At its north end, a short distance north of the property boundary, the Northwest Zone syncline plunges moderately to the south, and at its south end, near the road between the camp and Bisha Main, it plunges gently to the north (Figure 7-6). In the Tabakin Range, a crude dome is outlined by relatively continuous exposure of interbedded felsic and mafic metavolcanic rocks. Although the constituent folds in the range are tight and complex, they yield an overall anticlinal geometry, elongate parallel to the axial trace, and with a northerly plunge on the north and a southerly plunge near the south end of the range. Other culminations have been inferred, in part, from the distribution of mappable units, the airborne geophysical data, smaller-scale structures (e.g., the syncline on the northern side of Guardian hill), and from fold hinges inferred from bedding-cleavage relationships or reversals in bed dips all suggest that the structural style exhibited in the dome and basin structures predominates throughout the area mapped.

The largest-scale fold structure apparent on Figure 7-6 is the north-plunging antiform outlined by the contact of the Bisha gabbroic complex. It has a wavelength of 10 or 15 km and a probable amplitude of up to several kilometres(?). Folds such as the Bisha and Northwest Zone synclines, and the Tabakin Range dome or anticline, appear to be of an order of magnitude smaller, with wavelengths of up to a kilometre or so, and amplitudes of hundreds of metres. The distance along the trend of these folds between adjacent culminations and depressions appears to be somewhat greater than their amplitude, perhaps on the order of 2 km or more, this is consistent with the appressed nature of the folds. Several(?) basins and domes between the Bisha and Northwest Zone synclines, and to the east of the Tabakin Range structure, appear to have similar dimensions. Folds are, in general, much less evident on the outcrop-scale, although this is most likely a consequence of the lack of compositional contrast within the stratigraphy, to the lack of well-developed layering in many of the poorly-stratified and thick-bedded to massive tuffaceous rocks, as well as to the common strong overprint of foliation.

Although many individual folds at Bisha display a clear sense of structural vergence, the overall pattern of structural vergence remains to be clearly established, although the general younging direction of stratigraphy toward the west certainly suggests that the direction of tectonic transport is in that direction.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-23
	October 2004

Folds at Bisha are likely en-echelon in style, with one fold, or fold pair, terminating or relaying into another fold or fold pair–this may well be the case at Bisha Main, where hinges of synthetic folds on the western limb of Bisha syncline apparently end along trend to the north-northeast. It is likely that these folds may pass into, or over, a culmination, which plunges southerly on the southwest and northerly on the northeast. It is possible that the plunge reversals may be related to cross-folding, but if they are, then there are essentially no related fabrics or outcrop-scale structures related to another folding event, whether it is later or earlier. Instead, the relaying of en echelon folds or fold pairs, one to another, is the interpreted structural style, with the plunge reversals and commonly tightly appressed folds possibly exacerbated by indentation of the more competent “buttress” of gabbroic rocks of the Bisha complex, which lie to the east and southeast. This view is supported by the parallelism of the axial surface traces (and axial surfaces?) of property-scale folds with the pervasive foliation, and, in general, with the contact of the Bisha gabbroic complex. Together with the very tight folding, this suggests that the folds and strong foliation were formed during one major phase of contractional deformation, with the gabbroic complex perhaps acting as an “indenter,” perhaps most active toward the later stages of a somewhat prolonged(?) phase of crudely coaxial deformation.

Rock Fabrics

S1 Foliation. A well-defined S1 foliation is the most notable outcrop- and hand specimen-scale structural feature of the rocks at Bisha. And, with the exception of locally well-developed joint sets, it is typically the only fabric present in the generally poorly stratified rocks. The S1 foliation is more or less pervasive, and with the exception of the most competent of lithologies or locally within the noses of more open folds; it is rare that it is not well expressed. It is typically spaced on the millimetre- and more rarely on the centimetre-scale (in the most competent lithologies), and is defined in the predominantly felsic metavolcanic rocks by the presence of relatively abundant fine-grained white mica. Foliation intensities do not appear to vary greatly across the area mapped, and no areas with relatively high strain gradients, such as discrete ductile shear zones, were apparent in the field. The foliation appears to have a strong flattening component, but there also appears to be a significant component of stretching involved in its development (see section on Linear fabrics below). In general, the foliation is steep and strikes to the north-northeast, with more northerly and north-northwesterly trends becoming dominant to the east and north (Figure 7-6). These changes in foliation orientation, which in general are into parallelism with the contact of the Bisha gabbroic complex, imply that the competent mafic intrusive rocks have exerted some form of structural control.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-24
	October 2004

In the southeast, foliation trends are somewhat more disrupted, with east-northeast trends more common, and very locally, the foliation is folded. Where the foliation is folded, outcrop is limited, and the suggestion, from the proximity of a well-defined and lengthy topographic lineament, as well as from the lack of second-phase fabrics, is that the foliation has been disrupted (drag-folded?) by movement along nearby late-stage brittle faults.

S2 Fabric. As noted above, the S1 foliation is typically the only fabric developed in the rocks at Bisha. Locally, however, and particularly in the hinge zones of map-scale folds, an irregularly developed, commonly centimetre-spaced second phase (S2) foliation, best described as a spaced cleavage, overprints the S1 foliation. It is almost invariably at a small angle to the S1 foliation, and speculatively, it is believed to have been developed during continued tightening in the hinges of the folds.

S3 Fabric. In the area in the immediate vicinity of Bisha Main, a northwest trending, steeply southwest or northeast dipping crenulation cleavage locally cuts the well-developed S1 fabric in dacitic to rhyodacitic ash to fine lapilli tuff. Unlike the S2 fabric, this fabric is at a very high angle to the S1 fabric, and it is even more widely-spaced, with crenulation cleavage domains being as closely-spaced as several centimetres, but only very locally. Because of these differences, because the crenulation cleavage is much less widespread than are the S2-type fabrics, and because it was not observed in relation to the S2 fabric, it is referred to as S3.

Linear fabrics

Although the data are limited (n=22), stretching lineations are not uncommon within the rocks at Bisha. They are particularly notable within the coarser fragmental deposits, particularly medium to coarse lapilli tuff, and stretches in the range of 3 or 4:1, up to approximately 10:1, are not uncommon. In general, the lineation appears to mirror the plunges of the folds (e.g., southerly plunge in the vicinity of the Bisha syncline), with both gentle to moderate south-southeast plunges and moderate northerly plunges apparent in the data.

Faults and Lineaments

The stratigraphy and principal tectonic fabrics at Bisha have been disrupted at least locally by late-stage brittle faults. Because of the relatively poor exposure in the area, these are expressed in the main as well-developed topographic lineaments. The most prominent of these faults, such as the one 2.5-3.0 km east of the main deposits, are north trending, with the apparent offset of the Bisha Gabbroic contact suggesting that it has up to several kilometres of sinistral displacement. Many other well-developed lineaments are also present, but whether or not these features are faults remains uncertain (see Figures 7-6 and 7-7).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-25
	October 2004

7.4.4

Metamorphism and Weathering

Metamorphism

Nacfa Terrane greenstone belt rocks such as the volcanic and sedimentary units at the Bisha Property exhibit upper greenschist to lower amphibolite facies metamorphism. The presence of chlorite, fine grained amphibole, and local garnet in the mafic rocks supports that the grade of metamorphism has been reached (Greig, 2004).

Weathering

Weathering and laterite development from the beginning of the Cretaceous to the end of the Neogene have resulted in a locally deeply weathered terrain (Tardy et. al., 1988). However, Chisholm et. al. (2003) noted that the weathering profile is not similar to the laterite terrains of West Africa and Australia as the typical laterite-saprolite leaching sequences are absent.

The development of a gossan zone over massive sulphide zones is common to Bisha and several other deposits in the region (i.e. Debarwa, Adi Nefas; see Table 7-2). The gossans at Bisha occur as small hills with abundant black to brown pebbles and boulders composed of nearly 100% hematite and limonite (Photos A-2 to 6 in Appendix A). The gossan and boulders are the surface expression of deeply weathered (typically several tens of metres) massive sulphide horizons. Lines of boulders within the gossan are likely to be in-place weathered stratigraphic horizons. The gossans are locally traceable for well over 100 metres, and at the Bisha Gossan Zone, they outline the nose of the Bisha syncline (Greig, 2004).

The oxidation of the massive sulphides generated strong acid solutions that have progressively destroyed the sulphides and host rock. A horizon of extremely acid leached material has developed between the oxide and supergene/primary zones. The acid zone is typified by the “SOAP” unit intersected in drilling, which was originally assumed to be a type of footwall alteration and was termed “ALT” in logs. The unit was hypothesized to appear as a window in the core of an exposed syncline or anticline. SOAP is now interpreted to be composed of remnant clay and silica after exposure of the massive sulphides and host rock to an extreme acid environment as the sulphides became oxidized. On the northwestern limit of the exposed SOAP (at 1715932N, 339275E UTM) are two very small sub-crops of a rock type that are very well indurated, porous and blood red colour (hematite-stained) on fresh surface, which appears to be a sub-type of the SOAP rock category.

Further discussion of the oxide/gossan unit related to the Bisha Deposit is provided in Section 9.0.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 7-26
	October 2004

8.0

DEPOSIT TYPES

Bisha is a large precious and base metal-rich volcanogenic massive sulphide (VMS) deposit. Pertinent deposit model types would be Noranda/Kuroko (Franklin et. al., 1981) or bimodal-siliciclastic VMS deposits (Barrie, 2004). The Matagami Deposit in the Matagami VMS District in Quebec is a relevant and comparable deposit given the size (25 Mt), host rocks, proximity to a mafic complex, and several other features (Barrie, 2004).

8.1

Noranda/Kuroko VMS Deposit Model

Noranda/Kuroko style volcanogenic massive sulphide deposits are noted for their high grade polymetallic nature, associated precious metal content, moderate to large tonnages, and occurrence within districts of multiple lenses or horizons. Key characteristics (after Höy, 2004) of Noranda/Kuroko style volcanogenic massive sulphide deposits are:

• Marine volcanism, formed during period of felsic volcanism in an andesite or basalt dominated succession.
  

• Associated with faults, grabens, and prominent fractures.
  

• Associated with felsic or intermediate (or both) volcanic rock including epiclastics.
  

• Polymetallic (copper, lead, zinc plus gold and silver) massive sulphide deposits.
  

• Massive to well-layered sulphides, sedimentary textures.
  

• Quartz, chlorite, sericite alteration near core to clay, albite, carbonate minerals further out.
  

• One or more lenses within felsic volcanic rocks in a calk-alkaline bimodal arc succession.
  

• Cu-rich base, Pb-Zn rich top.
  

• Low-grade stockwork zones underlie lenses.
  

• Barite and chert layers, lateral gradation into chert horizons.
  

Each of these features is present at Bisha with the exception of the host volcanic rock geochemistry, which is subalkaline (Greig, 2004).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 8-1
	October 2004

Figure 8.1
Kuroko Style VMS Deposit Model¹⁵

Franklin et. al. (1981) presented a schematic section of the Kuroko volcanogenic massive sulphide deposit model (Figure 8-1). The model is simple relative to the geologic model of Bisha and the deformation, near-surface oxidation, and regional metamorphism at Bisha could easily have masked the alteration pattern and stockwork zone.

Bisha has an overall deposit size combined Indicated mineral resources of 22.7 Mt (Oxide, Supergene and Primary Zones) and an additional Inferred resource of 5.8 Mt (Oxide, Supergene and Primary Zones; see Section 17.0). The deposit is therefore larger than most of the typical Kuroko VMS deposits based on grade-tonnage models by Singer and Mosier (USGS, 1983; Figure 8-2A). The precious metal and base grades of the Bisha Primary Zone mineralization are in the higher percentiles for the grade-tonnage models (Figure 8-2B and C – base metal grade-tonnage models are not shown).

¹⁵  Modified from Franklin et. al., 1981 by Singer, 1986.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 8-2
	October 2004

Figure 8-2
Kuroko Style VMS Grade and Tonnage Model (Singer and Mosier, 1986)

8.2

Bimodal Siliciclastic VMS Deposit Model

Barrie and Hannington (1999) have proposed a five-part classification system for VMS deposits based on host rock composition. Two of the types potentially describe the Bisha Deposit: Bimodal Siliciclastic; or Mafic Siliciclastic. Barrie (2004) visited the Bisha Property and concluded that the bimodal siliciclastic model was most appropriate. Mapping by Greig (2004) also indicated that the host rock is principally felsic volcanic rock (variably altered felsic lapilli, lapilli ash tuffs, crystal tuffs and minor felsic dykes).

Many characteristics of the Kuroko VMS deposit model also apply to the Bimodal Siliciclastic VMS model. Bimodal Siliciclastic deposits form in lithologic sequences composed of roughly equal proportions of volcanic and siliciclastic rocks. Typically felsic volcanic rocks are more abundant than mafic rocks and are calc-alkalic in composition, while mafic rocks are of tholeiitic composition. Deposits are typically of Phanerozoic age and are typified by the deposits of the Iberian Pyrite Belt and the Bathurst camp of New Brunswick. Barrie (2004) considers the Bisha Deposit to be similar to those of the Iberian Pyrite Belt.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 8-3
	October 2004

Figure 8-3
Bisha Bimodal Siliciclastic VMS Model Schematic

Barrie (2004) developed a VMS model for the Bisha Deposit incorporating the local features such as the Bisha Gabbroic Complex and dominantly felsic and siliciclastic host rocks (Figure 8-3).

Bimodal siliciclastic deposits represent the largest VMS tonnage. However, on average they have the lowest Cu (1%) and the highest Pb (1.8%) metal content of the five deposit types (Barrie and Hannington, 1999), while also having relatively high Zn (4%), Ag (90 g/t) and low Au (1 g/t) contents.

Franklin (1998) described deposits of the Iberian Pyrite Belt and noted that they are characterized by great lateral continuity of ore as well as lack of extensive alteration. Both characteristics appear to have relevance to the Bisha Deposit. The Iberian Pyrite Belt deposits range from small lenses of a few million tonnes to very large bodies (over 100 Mt).

Either VMS deposit model is applicable to the Bisha Deposit.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 8-4
	October 2004

9.0

MINERALIZATION

9.1

Introduction

VMS mineral deposits on the Bisha Property include the Main Zone and the Northwest Zone (Figures 4-3 or 5-1), both of which have been drilled and have intervals of massive sulphides hosted within felsic volcaniclastic rocks. The Harena Area and NW Barite Showing have also been mapped, sampled and subjected to geophysical surveys but are not yet drilled. The focus of this section of the report is the mineralization of the Bisha Main Zone.

Four principal zones of mineralization within the Bisha Main Zone include: (1) a near-surface oxide/gossan; (2) a horizon that has been subjected to extreme acidification¹⁶ (acidified); (3) a supergene copper-enriched horizon; and (4) a primary massive sulphide horizon (Figure 9-1).

Figure 9-1
Isometric View of the Bisha Deposit Facing West

¹⁶ Acidification of the massive sulphides and host rock results remnant clay and silica, which is logged as ACID or SOAP rock codes.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 9-1
	October 2004

9.2

Host Rock

Characteristics of the host units to the Main Zone mineralization include:

• Precious metal rich (Au, Ag), base metal rich (Cu, Zn, Pb) massive sulphide lenses hosted by a bimodal sequence of weakly stratified, predominantly tuffaceous metavolcanic rocks (Nacfa Terrane greenstone belt).
  

• Host rocks are felsic lithologies (variably altered felsic lapilli and lapilli ash tuffs, crystal tuffs and minor felsic dykes), which form the hanging wall stratigraphy and predominate overall.
  

• Sub-alkaline (Greig, 2004) geochemistry of the volcanic rocks.
  

Earlier logging based on visual observations described the majority of the host rock as mafic (Chisholm et. al., 2003) and yet the most recent work and geochemistry supports felsic compositions (Barrie, 2004; Greig, 2004).

9.3

Deposit Dimensions and Morphology

The Main Zone is a 1.2 km long, narrow massive sulphide lens and is oriented north-south (Figure 9-2). The thickness of the lens is variable from 0 to 70 m but the deposit is deformed and exhibits thickening at the fold hinge and limb attenuation, which distorts original dimensions (see Figures G-1 to 5 in Appendix G).

Drill intersections encountered mineralization to a depth of 380 m but portions of the deposit only extend to depths of 70 m (i.e. east limb on section 1715500; see Figure G-1 in Appendix G; north portion of the deposit in Figure 9-1).

The tightly, complexly folded nature of the deposit and host stratigraphy is evidenced in Figures 7-7 and 9-2 (and also in Figures G-1 to 5 in Appendix G). Thickening at the fold hinges and exposure of a possible hinge of a south-plunging syncline are best noted between 1716000 to 1716300 N and 339200 E and 339500E (Figure 9-2).

A fault with a northwest strike is interpreted to have displaced the Main Zone upwards (displaced northeast side up) at 1716100 N (see Figure G-5; Greig, 2004). Another northwest trending fault has been postulated to occur at 1715400 N where the western lens (or limb) disappears from the sections and this may be due to vertical displacement of the deposit.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 9-2
	October 2004

Figure 9-2
Drill Hole Location and Bisha Main Zone Outline

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 9-3
	October 2004

The structural history of the deposit is not complete and differing interpretations are possible. For example: on section 1715300 North (Figure G-1), the two mineralized horizons could be two parallel lenses or the same horizon of opposing limbs of a fold. Another example is on section 1716100 North (Figure G-5), which could represent the keels of two horizons, which rapidly terminate or a single horizon deformed into a syncline-anticline-syncline combination. Nevsun considers that these represent two separate mineralized horizons.

The south end of the Main Zone plunges very rapidly to the south and drilling has been completed without success for 50 m south of the last mineralized intercept. However, the structural complexity (i.e. a fold or a fault) may have displaced the zone downwards and this area may still be considered to be “open” but only at depth. Nevsun has prepared a structural interpretation that warrants additional drilling (pers. comm. Nielsen, 2004).

The north end of the Main Zone appears to be abruptly terminated due the exposure and erosion of the keel of the Bisha Syncline (See Figure G-5 in Appendix G).

The deposit remains open down dip in several portions of the deposit. Extensions at depth would add primary sulphide mineralization.

Further opportunities exist along the limbs as noted in the structural interpretations on the deposit-scale geological map (Figure 7-7). The Northwest Zone, located approximately 4 km north of the Main Zone, is interpreted by Greig (2004) to be another exposure of the same mineralized horizon in a syncline. Mineralization was intersected in 8 of the 14 holes drilled in the Northwest Zone. One of the longest intervals is from hole B-066 with a 47.5 m core length interval averaging 1.32 g/t Au, 14.96 g/t Ag, 1.52% Cu, 0.01% Pb, and 0.04% Zn (pers. comm. Nielsen, 2004).

Metal zoning within the massive sulphide appears to indicate an upward transition from Cu-rich to Zn-rich to barren pyrite and confirms the interpretation that the sequence is “right-way-up” (west-facing).

9.4

Oxide Zone

The Main Zone is typified by the exposure of the Main Gossan, which is the only surface expression of the deposit (Photos A-7 and 8 in Appendix A). The gossan is a large mound of red-brown oxide material ranging from fine sand to dense cobbles and boulders distributed randomly or as groups or possible remnants of stratigraphic “horizons”. The boulders and cobbles are usually extremely siliceous. Much of the material is massive iron oxide (Photo A-18 in Appendix A). This portion of the Main Zone is flanked on the west by Guardian Hill (Photo A-5) and by another hill to the east (Photo A-6).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 9-4
	October 2004

The ferruginous to massive goethite-hematite-jarosite gossan is the remnant of surface oxidation of the massive sulphides. The depth of oxidation is variable but appears to be in the order of 30 to 35 m in outcrop areas. The unit has a high gold content, which is postulated to occur as precipitated micron-sized native gold that formed during the descent of oxidizing solutions. The relatively low base metal values (copper, zinc) are due to leaching during oxidation. The breccia domain with the Oxide Zone has gold values of 2.32 to 7.47 g/t Au (based on composite values at the 25^th to 75^th percentiles in Appendix G).

In some gossan outcrops there are banded white opaque quartz veins ranging up to 0.5 m in thickness and several metres in length. This material may be re-crystallized chert (Chisholm et. al., 2003).

Within the Oxide is a breccia unit that was separated out during geologic modelling and resource estimation due to higher gold grades than surrounding oxide material. The breccia occurs around (flanking) the gossan and appears to be a product of oxidation, laterite weathering, and desegregation of the original rock as opposed to being a structural feature. The unit is mostly quartz breccia or silicified fragments within oxidized material. The unit is poorly described and should be further examined during future studies.

9.5

Acid Zone

The extremely acidic nature of the oxidation of the massive sulphides caused the development of a highly leached “front”. Acid solutions leached the rock and the ACID or SOAP units are the very friable remnants consisting of mostly clay and silica (Photo A-19 and 21 in Appendix A). Although the transition from the Oxide to Acid Zone can be gradual (Photo A-23 in Appendix A); the change to the underlying Supergene Zone is rapid (Photo A-20 in Appendix A).

The thickness of the Acid horizon is variable ranging from 0.5 to 6 m, and averaging 3 m in thickness. This unit is mostly devoid of significant base metal mineralization but has high gold values. The gold grades between the 25^th to 75^th percentiles of the composites are 2.36 to 7.7 g/t Au (based on 3 m composites in Appendix G).

Due the crumbly nature of the SOAP unit, it can be difficult to obtain accurate measurements of the dry bulk density. Further work programs must include efforts to more accurately establish an acceptable value.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 9-5
	October 2004

9.6

Supergene Zone

The Supergene Zone is copper-enriched and occurs between 35 to 65 m depth. As in the supergene enrichment of porphyry deposits, oxidation of the massive sulphides caused the descending waters to become acidic and leach copper and other metals. The metals were deposited as chalcocite and covellite at the base of the Acid and Oxide Zones. Sooty secondary sulphides coat and replace primary sulphides.

The grades between the 25^th to 75^th percentiles of the composites for the Supergene Zone are 0.58 to 3.99 % for Cu and 0.37 to 0.92 g/t for Au (based on 3 m composites in Appendix G).

Other metals may have also been leached from upper levels and enriched in the Supergene Zone but they are not discussed here.

9.7

Primary and Primary Zn Zones

The primary sulphide mineralization at Bisha is below the Oxide, Acid and Supergene Zones, at a vertical depth of 60 to 70 m. Sulphide minerals are predominantly pyrite with some sphalerite and chalcopyrite. The grades of the Primary Zone between the 25^th to 75^th percentiles of the composites are 0.72 to 1.74% Zn, 0.26 to 0.93% Cu, and 0.36 to 0.76 g/t Au (based on 3 m composites in Appendix G).

Sphalerite appears to be more abundant at the south end of the deposit and a Primary Zn domain has been separated. The grades between the 25^th to 75^th percentiles of the composites for the Primary Zn Zone are 7.62 to 13.20% for Zn and 0.51 to 0.95% g/t for Au (based on 3 m composites in Appendix G).

Textures include semi-massive, massive, banded/laminated, minor folds, clasts, and disseminated sulphides within chloritized volcanics (Photos A-20, 22 in Appendix A).

The massive sulphides make the Primary Zone of the Bisha Deposit an excellent geophysical target (gravity, EM). Nevsun has completed ground and airborne surveys over the immediate Bisha Main area, the Northwest Zone, the Harena Area (9 km southwest) and the NW Barite Showing (6 km west). Several anomalies remain to be tested.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 9-6
	October 2004

9.8

Footwall Alteration

Footwall alteration is typically pervasive chloritic alteration of tuffs, which may extend for tens of metres below the massive sulphide unit. The position of the alteration zone supports the interpretation that the sequence is ‘right way up” (west-facing). Immediately below the massive sulphides there is also a thin but variable (<3m thick) zone of silicification and K-feldspar replacement (Chisholm et. al., 2003). This zone is more variable in intensity and thickness than the chlorite alteration and in some cases is entirely absent.

Barrie (2004) confirmed the presence of the chlorite alteration zone using the chlorite alteration index from ICP data. Barrie (2004) noted that the eastern lens (or limb) has footwall alteration along its entire length whereas the western lens (or limb) has alteration only at the north end of the deposit. This concurs with comments by several workers (pers. comm. Ansell, 2004; Chisholm et. al., 2003) that the possible hydrothermal fluid source or conduit may be at the north end of the deposit.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 9-7
	October 2004

10.0

EXPLORATION

10.1

Introduction

Exploration of the Bisha Property commenced in 1996 as prospecting by Ophir Ventures. Nevsun subsequently explored the Property in several programs of work over a seven-year period between 1998 and 2004 (Table 10-1). Exploration was suspended between 1999 until 2001 during the border war with Ethiopia.

In June 1999, the Bisha Prospecting License was converted to the Bisha Exploration License covering an area of 49 km². In 2003 the license area was expanded and now includes a total area of 322 km² (see Section 4.2).

In October 2002, Nevsun initiated a limited diamond drilling program at Bisha in order to test the combined geophysical and geochemical anomaly over the main gossan outcrop area. A summary of the significant assay intervals is provided in Table 10-2.

To follow up the successful results from the 2002 drilling, Nevsun completed two phases of diamond drilling in 2003. The Phase I work was completed between February and June and consisted of 48 holes totalling 6,724.76 m plus mapping, sampling, trenching, geophysics (airborne and ground), mapping, metallurgical testing, and bulk density measurements. The Phase II work was conducted from September until December and consisted of 92 core holes totalling 11,894.50 m. Additional work conducted during this program included geophysics, geochemical sampling, metallurgical testing, petrographic work, and bulk density measurements (Table 10-1).

Further diamond drill holes, RC holes, and combination holes totalling 31,285.60 m were completed during 2004. Additional exploration work completed during this program included geophysical surveys, mapping, geochemical sampling, petrographic work, bulk density measurements, geotechnical work, environmental and archaeological studies, and metallurgical testing (Table 10-1).

On September 2, 2004, Nevsun Resources Ltd. (NSU/TSX), Sanu Resources Ltd. (SNU/TSXV) and Sunridge Gold Corp (SGC/TSXV) each received a letter from the Minister of Energy and Mines for Eritrea instructing the companies to halt all mineral prospecting and exploration work and related activities in Eritrea until further notice. No reason was given for this instruction in the letter. Nevsun has temporarily suspended fieldwork in response to the stop work order.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-1
	October 2004

Table 10-1
Summary of Work Complete

Year	Phase	Company	Type of Work	Description
1996		Ophir Ventures	Regional Grassroots Exploration	Prospecting, mapping and sampling
1998		Nevsun	Property Evaluation	Property Examination and acquisition
1998		Nevsun	Property Grassroots Exploration	Reconnaissance scale geological mapping (1:50,000), geochemical stream sediment sampling
1999		Nevsun	Geophysical Surveys	Geophysical surveys - MaxMin horizontal loop EM and magnetometer
			Geological Mapping	Property scale (1:5,000)
			Geochemical Sampling	Soil sampling on three grid lines
2002		Nevsun	Drilling	6 diamond drill holes (B-01 to B-06) totalling 759.0 m
			Geological Mapping	Discovery outcrop area (1:1,000)
2003	I	Nevsun	Drilling	47 diamond drill holes (B-07 to B-53, B-02a) totalling 6722.6 m
			Trenching	36 trenches sampled and mapped
			Geophysical Surveys	Airborne EM and Magnetometer (325 sq km), Pulse EM and ground Magnetometer (73.5 line km), Gravimetric survey (40 km)
			Geological Mapping	Deposit scale (1:1,000), property scale (1:2,500) and regional scale (1:10,000)
			Geochemical Sampling	Stream sediment (165 samples), soil (39 samples), termite mound (115 samples), auger and pit (33 samples)
			Petrographic Study	11 thin sections by Vancouver Petrographics
			Metallurgical Testing	2 oxide samples, 2 copper supergene mineralization and 2 primary mineralization samples
			Bulk Density	260 samples determined on site, 44 samples sent to ALS Chemex for determination
2003	II	Nevsun	Drilling	93 diamond drill holes (B-54 to B-146, & deepen B-40) totalling 11,750.8
			Drilling	2 air blast holes for water wells completed by Eritrean Drilling
			Geophysical Surveys	Pulse EM, Horizontal loop EM (151 line km), Gravimetric survey (107.6 km)
			Geochemical Sampling	pH soil survey, soil sampling, Whole Rock (REE), regional prospecting
			Metallurgical Testing	2 oxide samples and 2 copper supergene mineralization samples at PRA in Vancouver, some minor work at Kappes Cassidy in Nevada
			Petrographic Study	13 thin sections by Vancouver Petrographics
			Bulk Density	611 samples determined on site, 68 samples sent to ALS Chemex for determination
2004	I	Nevsun	Drilling	163 diamond drill holes (B-147 to B-309) totalling 28,879.50 m
			Drilling	42 reverse circulation drill holes (BRC-01 to BRC-42) totalling 2,097.3 m
			Drilling	9 combination reverse circulation with diamond drill core holes tails (BRCD-26,27, 32 to 34, 37,38, 41 and 42) totalling 308.70 m
			Drilling	15 reverse circulation holes for water wells totalling 768 m.
			Geophysical Surveys	Gravimetric survey, 65.2 line km for a total to date of 215 line km
			Geological Mapping	Deposit scale (1:1,000) and regional prospecting
			Geochemical Sampling	Soil sampling (151.94 line km), Whole Rock (REE), prospecting
			Petrographic Study	16 thin sections, 2 polished sections
			Bulk Density	311 samples determined on site, 697 samples sent to ALS Chemex for determination
			Geotechnical Work	all drill core oriented
			Environmental	base line study implemented
			Metallurgical Testing	2 primary sulphide samples tested at PRA in Vancouver
			Hydrological	studies commenced
			Archaeological	studies commenced
			Physical Properties Tests	On selected core samples of massive sulphide by JVX Geophysics

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-2
	October 2004

Table 10-2
Significant Assay Intervals from 2002 Drill Program

Hole #	From	To	Interval (m)	Au g/t	Ag g/t	Cu %	Pb %	Zn %
B-2	29.00	66.00	37.00	0.02	0.65	0.93	0.00	0.00
B-3	4.35	14.50	10.15	1.96	1.72	0.07	0.06	0.04
B-3	19.0	28.96	9.96	10.24	44.80	0.07	1.95	0.04
B-3	134.95	172.00	37.05	0.99	24.94	0.97	0.04	1.92
incl	134.95	155.00	20.05	1.46	40.4	1.52	0.06	3.09
B-4	48.77	56.39	7.62	5.44	88.53	0.10	0.92	0.03
B-4	56.39	101.20	44.81	0.87	27.13	3.92	0.11	0.34
B-5	37.50	45.72	8.22	8.53	693.35	0.06	9.65	0.01
B-5	45.72	57.00	11.28	16.52	475.32	3.62	8.28	0.02

10.2

Coordinates and Datums

The coordinate system used for all data collection and surveying is the Universal Transverse Mercator (UTM) system, Zone 37 and geographic coordinates in WGS84¹⁷ (World Geodetic System 1984).

10.3

Topography and Grid Survey Control

During the 1999 exploration program, Nevsun established a local grid (not based on UTM coordinates) over the gossan area with a base line 5.9 km in length oriented at an azimuth of 010° from magnetic north. Individual lines were usually spaced 200 m apart and were of variable length.

The local grid constructed at the beginning of the 2003 program conforms to the UTM coordinate system with a baseline oriented at 0° and cross-lines oriented 090°. Cross-lines were usually spaced 100 m apart, except over the Bisha Main Zone where gridlines were at spaced 25 m apart for drilling.

The SW grid, that covers the Harena Area, has a baseline trending 045° with 100 m spaced gridlines. The Okreb area had numerous grids with varying baselines established over it in 1999 but none are now in existence. These two areas can be seen on Figure 10-1 and are discussed in Section 10.4 relating to mapping of the grid areas.

A small network of topographic control points were created both on and off the Bisha Property. All of the control points were established by post-processed static GPS observations. A summary of the key control points with locations given in latitude and longitude is shown in Table 10-3. The control point locations have been converted to UTM coordinates for Zone 37 for use in surveys on the Property.

¹⁷ WGS84 is the datum to which all GPS positioning information is referred by virtue of being the reference system of GPS satellites. WGS84 is an earth-fixed Cartesian coordinate system.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-3
	October 2004

Table 10-3
Control Points in WGS84 (Geographic Coordinates)

Point	Latitude	Longitude	Height	Elevation
CAMP	N 15deg 32' 50.93540"	E 37deg 29' 33.43519"	558.0191m	557.5999m
HILL	N 15deg 31' 47.12984"	E 37deg 29' 36.01355"	571.9160m	571.5526m
CON1	N 15deg 30' 48.86642"	E 37deg 29' 49.40324"	574.2377m	573.9299m
Source: MWH Geo-Surveys Inc., 2004

10.4

Geological Mapping and Related Studies

Nevsun has carried out geological mapping during each exploration program as summarized in Table 10-4.

Table 10-4
Summary of Geological Mapping on Bisha Property

Year	Type of Mapping	Scale	Location
1998	Reconnaissance	1:50,000	Bisha region
1999	Property	1:5,000	Bisha Main area
2002	Deposit	1:1,000	Bisha Main gossan
2003	Regional	1:10,000	Bisha region
2003	Property	1:2,500	Bisha Property
2003	Deposit	1:1,000	Bisha Main gossan
2003	Property	1:2,500	Barite, Harena and SE zones
2004	Deposit	1:1,000	Bisha Main zone and NW zone

Approximately 75% of the Bisha Property is covered by alluvium of varying thickness. Outcrop exposure is restricted to outcrops along seasonal watercourses and areas of topographic relief above the gently northerly-sloping alluvial plain.

The locations of detailed geological mapping completed on the Property are shown on Figure 10-1. Mapping focused on areas of gossan outcrops such as the Bisha Gossan Zone and the gossan at the Harena Area (Figure 10-1). Mapping of the SE area and the NW Barite Showing area (Figure 10-1) was completed to investigate geophysical and geochemical anomalies in those areas.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-4
	October 2004

Figure 10-1
Bisha Exploration License Geological Mapping Location Map

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-5
	October 2004

Several persons have mapped portions of the property (Nutt, Childe, Greig, Deliele, Chisholm, Ouattara) or provided more regional coverage (Woldu). The lack of outcrop, lack of marker horizons, and the structural complexity has hindered comprehensive geological interpretations. Landsat information and geophysical data aided the development of interpretations.

In 2004, C. Greig (and Taiga Consultants Ltd.) prepared a compiled property geology map (Figure 7-6) and developed a complementary deposit-scale geology map of the area from the Bisha Main Zone to the Northwest Zone (Figure 7-7).

AMEC completed a brief review of the geological interpretations in the immediate area of the Bisha Main Zone and found them to be consistent with the field observations (i.e. exposures on the Main Gossan, Fereketta River bed, Guardian Hill, Conical Hill, Northwest Zone).

10.5

Remote Sensing and Satellite Imagery

Nevsun prepared Landsat images of the area in 1998. The image (PIX format file) was re-georeferenced using available topographic maps and then output as an ECW (ER Mapper) format file. At the same time the digital, 1:100,000 scale Russian topographic maps were obtained from East View Cartographic out of the USA.

Geoanalytic of Calgary scanned, registered and compiled the Eritrean topographic maps and the Landsat image. ECW (ER Mapper) files of the topography & LANDSAT files were prepared. TAB files of the Bisha Exploration License showing property boundaries were added. Using the Landsat image with a translucent (30%) topographic overlay, a preliminary interpretation was made of the immediate Bisha occurrence area. The interpretation focused on structural features that were subsequently plotted on the geology map as well as the preliminary gravity maps.

During 2003, Earth Resource Surveys Inc. (ERSI) based in Vancouver, completed a remote sensing investigation for the Bisha Property and western Eritrea. The survey mapped alteration types and interpreted major structural features using different Landsat bands. The study highlighted alteration and structural trends (Chisholm et. al., 2003).

10.6

Geochemistry

10.6.1

Stream Sediment Sampling

Nevsun carried out stream sediment sampling in 1998 covering an area of 100 km² including the Bisha Prospecting License.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-6
	October 2004

Figure 10-2
1998 and 2003 Stream Sediment Survey Results

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-7
	October 2004

During the 2003 Phase I exploration program a stream sediment survey was designed, implemented and completed by M. Mercier, an independent geochemical consultant with Analytical Solutions (Toronto). The survey included a total of 165 samples collected at an approximate density of one sample per square kilometre (Figure 10-2).

The stream sediment surveys were considered to be an effective method of delineating areas with potential for base and precious metal mineralization. The main anomalous areas for copper, lead, zinc and gold based on the combined 1998 and 2003 results are: Okreb area (eastern portion of the Bisha Exploration License), the Bisha Main Zone southwards towards the Harena Area, and the NW Barite Showing area (Figure 10-2).

AMEC has not observed the stream sediment sampling practices. The sampling procedures as described are reasonable and are within standard industry practices.

10.6.2

Soil Geochemical Sampling

In 1999, soil samples were collected over the Bisha gossan outcrop (Nevsun 2003). Figure 10-3 clearly outlines the anomalous nature of the silver, copper, lead and zinc over the outcrop area. The samples were not analyzed for gold.

Figure 10-3
Soil Sample Results Over the Bisha Gossan Zone

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-8
	October 2004

Soil geochemical sampling by Mercier (2003) included orientation soil, termite mound, auger and pit sampling as described below:

• Soil and auger survey with a total of 39 samples (23 soil samples and 13 auger samples) to test the auger as a geochemical tool and also to verify the geochemical response of different regolith units.
  

• Pit samples with a total of 33 samples along a single line to make a short investigation of the regolith units.
  

• Termite mound survey with a total of 115 samples (including 8 auger test samples).
  

• Size fraction analysis of four samples to help with the regolith investigation.
  

More soil sampling was completed in 2004 with 133.62 line kilometres completed over the Bisha Main, Harena and NW Barite Showing areas. To date, a total of 6,287 soil samples or 151.94 line kilometres of sampling have been collected on the Bisha Property. Figure 10-4 shows the areas of soil and geochemical sampling. The majority of the sampling was completed in 2003 and 2004 on gridlines with coordinates conforming to UTM, Zone 37. Soil sampling was used to investigate geophysical anomalies, often in areas with minimal or no outcrop. Apparently the alluvial cover has not precluded the usefulness of the soil sampling in defining anomalies over previous geophysical targets and known mineralization.

AMEC has not observed the soil or other geochemical sampling practices. The procedures as described appear to be reasonable and within standard industry practices.

PH soil geochemical surveys were conducted over the known Bisha Gossan to test the theory that a pH measurement can identify the change in pH related to the presence of massive sulphides (and related generation of acid conditions related to oxidation of the sulphides) even below alluvial cover. Nevsun found that the pH technique works very well in the delineation of known sulphide mineralization in alluvial covered areas and considered that it may be used to define new targets. Unfortunately the survey reacts to a wide variety of types of underlying chemical differences and thus produces a large number of anomalies that need to be prioritized.

AMEC did not review the method and cannot comment on the relative merits for further pH geochemical surveys.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-9
	October 2004

Figure 10-4
Geochemical Sampling Areas on the Bisha Property

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-10
	October 2004

10.7

Trenching

A series of 36 trenches were excavated with a Komatsu WB93R utility tractor equipped with a backhoe over various parts of the Bisha Main and Northwest Zones. Details of the trenching are provided in Section 12.6.

The trenches were completed on the Bisha Gossan Zone and Northwest Zone. Rock samples returned elevated base metal and precious metal results, as expected. The trench data was not used in geological modelling or resource estimation.

10.8

Ground Geophysics

10.8.1

Electromagnetic (EM)

Horizontal loop (HLEM) surveys were completed in 1999 by a South African based consulting group, Integrated Geophysical Surveys on the portions of the property. The 1999 grid was surveyed with a MaxMin II HLEM. The EM survey used frequencies of 444 and 1,777 Hz and a cable length of 150 m. Further work was completed in 2003 by Geophysique TCM of Val d’Or, Quebec, Canada. The 2003 survey included 151 km of Horizontal Loop EM (HLEM) surveys on the UTM grid at 100 m spacing.

Pulse EM surveys were contracted to Crone Geophysics (Crone) of Mississauga, Ontario, in 2003. A total of 43 lines over 13 loops consisting of 73.5 line km of survey were completed as shown in Figure 10-5.

Crone attempted downhole EM surveys on 3 drill holes. Two of the holes were caved but hole B-104 was probed successfully.

Most of the electromagnetic surveys were successful in delineating mineralization and other features (i.e. structures, lithologic contrasts, etc.; pers. comm. Ansell, 2004).

10.8.2

Magnetometer

Magnetometer surveys were completed on the portions of the property in 1999 and 2003.

In 1999, Integrated Geophysical Surveys completed the magnetometer survey on the grid established over the Bisha Main Zone. Eritrean technicians were trained by the South African consultants to carry out the survey.

In 2003, a magnetometer survey by Crone Geophysics was completed over the Bisha Main Zone covering a total of 43 lines covering 84.5 line km (Figure 10-5).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-11
	October 2004

Figure 10-5
Ground Geophysical Survey Compilation Map

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-12
	October 2004

The magnetometer surveys shows lithologic contrasts and anomalies over the Bisha Main Zone but were less distinct than the electromagnetic surveys. Chisholm et. al. (2003) observed north-northeast trending features that he suggested could represent a fault zone that transects the Bisha Deposit.

10.8.3

Induced Polarization (IP)

An Induced Polarization (IP) survey was completed on the Okreb area, in the northeast corner of the Bisha Exploration License. The survey was done on several grids with baseline at different orientations but generally trending northeast (see Figure 10-5). Cross-lines were completed at 200 m spacing along the baselines.

10.8.4

Gravity

Gravity surveys were carried out during the three consecutive exploration programs since 2003. MWH Geo-Surveys, Inc. of Reno, Nevada, used a LaCoste & Romberg Aliod 100X gravity meter to complete the work.

The dataset consists of 8,613 stations, including 8,043 stations collected on the Bisha grid, 407 stations collected on the Harena Area and 163 stations collected on the NW Barite Showing area (as shown in Figure 10-6). The stations were collected along UTM gridlines at 100 m line spacing and at 25 m station intervals. The survey over the Harena Area follows the grid in that area and has a 100 m or 200 m line spacing and 25 m station intervals.

The filtered residual gravity surveys provided good definition of the Bisha Main Zone massive sulphide mineralization.

10.9

Airborne Geophysics

In March 2003, a combined airborne EM and magnetometer fixed-wing survey was conducted over an area of approximately 325 km² corresponding to the entire Bisha Property. The survey was contracted to Fugro Airborne Surveys based in Ottawa, Canada.

A nominal line spacing of 100 meters was maintained in an east-west direction at an EM sensor altitude of approximately 73 m above ground level. A total of 4,052 line km of data was collected over the area. Fugro Airborne Surveys provided a report entitled “Logistics and Processing Report Airborne Magnetic and GEOTEM Survey, Bisha Area, Gash Barka District, Eritrea Job No.03427”. The report and accompanying maps provide all the technical aspects of the survey work.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-13
	October 2004

Figure 10-6
Gravity Coverage on Bisha Property

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-15
	October 2004

10.10

Mineralogical and Petrographic Studies

In 2003, 28 polished thin sections were collected from drill core and rock samples and sent to Vancouver Petrographics Ltd. for petrographic study. The thin section samples were collected from lithologies that were proving difficult to identify due to strong alteration and weathering (pers. comm. Ansell, 2004).

A further 18 drill core samples were selected and sent to Vancouver Petrographics for petrographic study in 2004. Sixteen samples were for thin section and two samples were for polished sections. The two polished section samples were taken from the primary massive sulphides representing both the zinc rich zone and the copper rich zone (see Appendix E for the mineralogy report on the two polished sections).

10.11

Bulk Density Determination

Refer to Section 11.2.7 for a detailed description of the bulk density determinations.

A total of 1,991 bulk density determinations (also referred to as specific gravity, “SG”, in parts of this report) on drill core have been completed by Nevsun and at ALS Chemex. Data was collected from a representative suite of all rock types, mineralization types and grade ranges.

10.12

Preliminary Metallurgical Studies

Refer to Section 16 for a description of the metallurgical testwork.

Metallurgical testwork included cyanidation of two high grade oxide gold samples from two drill holes in Bisha Main Zone; and flotation testwork on two samples of supergene copper mineralization.

10.13

Drilling

Refer to Section 11 for a detailed description of the drilling completed on the Bisha Property.

In October 2002, Nevsun completed a limited diamond drilling program at Bisha for a total of 810.90 m in order to test the geophysical and geochemical anomalies at the Bisha Gossan Zone outcrop area.

After the successful results from the 2002 drilling, two Phases of diamond drilling were completed in 2003. The Phase I work was completed between February and June and consisted of 48 holes totalling 6,724.76. The Phase II work was conducted from September until December and consisted of 92 holes totalling 11,894.50 m.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-16
	October 2004

Further core, RC, and combination drill holes were completed during 2004 totalling 31,285.60 m in 205 drill holes. All diamond drill holes completed in 2004 were orientated for geotechnical data collection.

10.14

Other Studies

Other studies completed or under way on the Bisha Property include: environmental baseline studies, archaeological investigations, and hydrological studies. As part of the hydrological study a series of seven monitoring water wells for 489.9 m were drilled on the periphery of the proposed open pit area and five production water wells totalling 278.0 m were drilled within the concession boundaries under the supervision of Klohn Crippen Consultants Ltd. (see Section 11.4 for a description of water well drilling).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 10-17
	October 2004

11.0

DRILLING

11.1

Introduction

Nevsun has conducted exploration activities on the Bisha Property since 1998. Core drilling (NQ and HQ diameter) and a small amount of RC drilling were completed on several zones in 2003, 2003 and 2004. Figures 9-2 and 11-1 show the surface trace of all drill holes in the Bisha Main Zone and South Zone. Drill hole collar information is provided in Table C-1 in Appendix C.

11.2

Diamond Drilling

Nevsun has completed a total of 50,715.76 m of drilling in 352 holes (see Table 11-1). Of this, 48,309.66 m was core drilling in 310 holes; 1814.4 m was completed in 33 RC holes; and 591.70 m was in 9 combination holes, which had RC at the top of each hole and core drilling in the bottom.

Table 11-1
Drill Hole Summary by Year and Type

Year	Phase	Range of Hole Numbers	Number of DDH Holes	Length of DDH (m)	Number of RC Holes	Length of RC (m)	Total Number of Holes	Total Length (m)
2002		B-001 to 6	6	810.90			6	810.90
2003	I	B-002a, 7 to 53	48	6,724.76			48	6,724.76
2003	II	B-054 to 146	93	11,894.50			93	11,894.50
2004		B-147 to 309	163	28,879.50			163	28,879.50
2004		BRC-001 to 40*			33	1,814.40	33	1,814.40
2004		BRCD-026 to 42*	9	308.80		282.90	9	591.70
TOTAL			319	48,618.46	33	2,097.30	352	50,715.76
* Not a continuous series

Drilling in late 2002 intercepted significant intervals of precious and base metal mineralization at Bisha below the Main Gossan. Nevsun mounted a major exploration effort in February, 2003 (Phase I) including trenching, geological mapping, geochemistry, geophysics (ground surveys, airborne EM and magnetometer survey) and drilling to define the mineralization. Further exploration and drilling programs were conducted in September 2003 (Phase 2) and January 2004.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-1
	October 2004

Figure 11-1
Drill Hole Location Map

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-2
	October 2004

The drilling of the first 17 holes¹⁸ for a total of 2,409.73 metres was begun in late 2002 and completed at the beginning of Phase I in 2003 using Kluane International Drilling, a contractor based in Vancouver, BC, Canada. Kluane used a “man-portable” drill rig powered by three water-cooled Kubota three cylinder diesel engines driving three hydraulic motors that feed a hydraulic head, wireline and mud tank mixer. The unit uses a 5’ long NTW core barrel (55.1mm diameter core), which was reduced in bad ground to a BTW sized (41.0 mm diameter core) 5’ or 10’ long core barrel.

The remaining 292 core holes for a total of 45,899.93 metres were completed during 2003 Phase I and Phase II and during 2004 Phase I using two Longyear 44 skid-mounted wire-line rigs (Photo A-12 in Appendix A), owned and operated by Boart Longyear, a contractor based in North Bay, Ontario, Canada. Each hole was collared with HQ core (63.5 mm diameter) until ground conditions necessitated a reduction to NQ sized core (47.6 mm diameter).

Nine combination holes were drilled (308.80 m core and 282.90 m RC) in 2004 using a universal drill rig (UDR-650-P35 combination drill; Photo A-15 in Appendix A) operated by Major Pontil Pty Ltd. (an Australian subsidiary of Major Drilling Inc.), based in Queensland, Australia.

A list of all drill holes completed on the Property is provided in Appendix C.

11.2.1

Collar Surveys

Nevsun placed a drill rod within cement at the collar of each hole to identify the hole location for all programs (Photo A-15 in Appendix A). The hole number was marked in the cement base. The diamond drill holes are identified by the prefix ‘B’, reverse circulation holes by the prefix ‘BRC’ and RC holes with diamond tails are denoted by ‘BRCD’.

Collar locations were surveyed by an independent contractor, MWH Geo-Surveys Inc. of Reno Nevada, U.S.A. using an Ashtech Z Xtreme dual frequency GPS receiver with real time kinematic positioning for sub-centimetre accuracy.

At the time of the AMEC site visit, 29 drill holes had not yet been surveyed but were subsequently completed at the end of drill program. The core and RC holes were checked by AMEC for obvious coordinate errors, swaps and extreme values. The reported coordinates were checked against a topographic map for 56 holes (19% of database) and no errors were noted for E or N. Problems noted for elevations were summarized in an AMEC memo (036-04.doc) and were resolved by Nevsun prior to completion of the drill hole database. The discrepancies required Nevsun to complete a comparison of all collars

¹⁸ All of the 2002 program and part of 2003 Phase I.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-3
	October 2004

to the DTM elevations at the collar positions. AMEC also requested a resurvey of 6 collars in the field (Photo A-15 in Appendix A).

11.2.2

Downhole Surveys

Nevsun used downhole survey instruments to collect the azimuth and inclination at specific depths of the diamond drill holes for programs by Boart Longyear and Major Pontil. Three types of survey method were used, these are: acid tests, Sperry-Sun Single-Shot and Reflex (Table 11-2). Most of the early drill holes have only collar surveys. Both the Sperry-Sun and Reflex units derive azimuth measurements using a magnet and are therefore subject to potential problems that can be caused by magnetic minerals. AMEC concluded that the rock units do not have an adverse affect on the downhole surveys.

Table 11-2
Drill Program Survey Methods

Drill Holes	Survey Method
B-01 to B-13, B-15, B-16 and B-02a	Collar survey only
B-14	Acid test
B-17 to B-53	Sperry-Sun
B-54 to B-309	Sperry-Sun or Reflex

The azimuth and inclination surveys of the initial 17 holes were collected at the collar using a Brunton compass and no downhole surveys were collected. One hole (B-14) had an acid test completed at the end of hole. Boart Longyear drillers used a Sperry-Sun instrument to take measurements at relatively wide intervals (20 m to 120 m) for holes B-17 to B-53. From holes B-54 to B-309, two instruments were employed by Boart Longyear, the Sperry-Sun and Reflex. The two instruments were divided between the drills and were used only on that machine unless a problem was noted with a measurement and confirmation of the reading required a check survey done by the other instrument. The measurements were taken at an initial 20 m depth down the drill holes and subsequently every 50 m thereafter unless hole conditions dictated otherwise.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-4
	October 2004

The measurements compiled by Nevsun were reviewed in detail by AMEC to find errors or omissions in the data. The exploratory data evaluation discovered an error in the value Nevsun used for magnetic declination, which is the correction to UTM north from magnetic north, used by Nevsun. Nevsun had used a declination of +4° for all corrections to UTM. The examination of the data concluded that the correction for magnetic declination must be changed. The actual declination was determined to be 2° 10’ for the project location at latitude 15° 31’ N and longitude 37° 30’ E as of the year 2004 and increasing 2’ east per year (Steve Manser, MWH Geo-Surveys Inc., pers. comm. 2004, www.geolab.nrcan.gc.ca/geomag/mirp_e.shtml).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-5
	October 2004

Figure 11-2
Average Azimuth and Dip Change Compared with Collar

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-6
	October 2004

Based on AMEC’s recommendations Nevsun agreed to a correction of -1° 30’ to all of the previous downhole measurements. The declination value now applied to all new measurements is +2° 30’.

Figure 11-2 (A and B) shows the average of the differences (deviations) in azimuth and dip between survey depths (i.e. between collar and first downhole survey, collar and second downhole survey, etc.) for holes grouped by the initial inclination of the hole for Sperry-Sun and Reflex instruments. The latter two graphs in Figure 11-2 (C and D) shows the deviations between surveys intervals. The initial deviation of the azimuth for all measurements is -0.40° and gradually increasing with depth. The Reflex instrument shows a larger initial deviation than the Sperry-Sun. The dip change for all measurements is negligible and gradually increases at 0.025° per m with depth. Sperry-Sun measurements show a greater deviation than the Reflex measurements.

The azimuth deviations gradually decrease with greater depths (Figure 11-2C) whereas the deviations in inclinations appear to increase until 200 m depth whereupon the deviations become less (Figure 11-2D).

Figure 11-3 shows the change in azimuth with the dip of the hole. The Sperry-Sun has a broader range of deviations at the –40° to –45° inclinations. The Reflex has a broader range of deviations for all inclinations.

Figure 11-3
Change in Azimuth with Dip

11.2.3

Logging

The core logging and storage facility include a large covered area for logging, handling, splitting and storing of the core within the camp perimeters (Photo A-16 and 17 in Appendix A). After the core arrives in camp, it is washed and meterage blocks are checked to ensure that no errors are present in the runs recorded during drilling.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-7
	October 2004

The logging system includes codes for key aspects of the geology and the style of mineralization. The mineralization is logged on hardcopy log forms. The logged information is later entered into an Excel spreadsheet by the same geologist that logged the hole. The key information is later geology information is later extracted from the digital logs in Excel and added to the drill hole database. Conversion to a standard database would assist in automatic presentation logs and allow for data filtering and pick lists to minimize opportunities for entry errors. AMEC also recommends adopting a system of double entry of the data.

The key data types captured during logging of the core are:

• Major lithology.
  

• Subsidiary lithology.
  

• Alteration.
  

• Mineralization.
  

• Structure.
  

• Sample intervals.
  

The geologist determines the sampling intervals and adheres to lithologic intervals. The sample intervals are identified by paper tags indicating a sample interval and felt pen marks on the core for the beginning and end of each sample. The geologist also marks the cut line for the core cutters to follow if there is potential for an apparent bias in mineralization. The cut lines are marked along the axis of the core perpendicular to the mineralized interval.

Observations of the core logging determined that a minimal number of persons were involved in the logging of core and continuity between logs appeared to be reasonable. An updated list of permissible logging codes should be produced given the increased geochemical knowledge of lithologies and other codes need to be standardized (i.e. litho, mod, mod 2 codes are not uniformly applied) to improve the development of the geological model. Plots of sections reviewed at site did not have the most recent holes plotted and interpretations were not current on all sections. The interpretations should be continually maintained during the drill program, and not prepared after the program is completed.

11.2.4

Photography

Digital photographs are taken of all drill cores; see Photos A-31 to 35 in Appendix A for a similar format of photos.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-8
	October 2004

11.2.5

Recoveries

Core recovery and RQD are measured at the drill rig as the core is placed in the core boxes (Photo A-13 in Appendix A). Nevsun’s Eritrean geology staff performs this work. The methodology used for measuring recovery was reviewed by AMEC and is standard industry practice. The data captured includes:

• Block interval.
  

• Drill run (m).
  

• Measured length (m).
  

• Calculated recovery (%).
  

• RQD measured length (m).
  

• Calculated RQD (%).
  

No recovery measurements were collected during the first drilling program (2002) or until hole B-13 at the beginning of Phase I drilling in 2003 (see Table 11-3). Average recoveries in the 2003 Phase I program, for the holes that had data collected was 88%. The recoveries for the Phase II program averaged 85%. The 2004 program for core holes drilled by Boart Longyear (163 holes) averaged 94% recovery, while the core portion of the mixed core/RC holes drilled by Major Pontil Pty. Ltd. (9 holes) had 88% recovery.

Table 11-3
Recovery by Drill Program

Year	Phase	Number of DDH Holes	Number of Runs Measured	Theoretical Length Being Measured (m)	Total Measured Core Length (m)	Average Recovery (%)
2002		6	0.00	-	-
2003	I	48	2298.00	5344.69	4715.30	88%
2003	II	93	3838.00	11537.50	9801.10	85%
2004		163	10517.00	28790.70	26933.09	94%
2004		9	163.00	281.60	247.22	88%
TOTAL		319	16816.00	45954.49	41696.72	91%

A total of 16,816 core runs were measured for the core recovery and a global average value was 91% (for 94.5% of the core drilled on the Property)

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-9
	October 2004

Given the nature of the Bisha mineralization, there are significant differences in recovery for the rock types due to the changes in lithology, alteration, and rock hardness. These factors result in poorer recovery for the near-surface mineralization of the oxide material and excellent recovery in the competent supergene and primary massive sulphide mineralization (see Table 11-4). High core loss was also common at the abrupt change from hard and relatively competent oxide material to the extremely soft SOAP lithologic unit.

AMEC investigated the potential for relationships between recovery and grades and no bias could be demonstrated. Domains such as the Fe-oxide Zone or the acidified SOAP unit typically have elevated precious metals even for intervals with high recoveries therefore a direct relationship to recovery cannot be justified. Recovery-grade relationships must continue to be investigated in future resource estimates.

Table 11-4
Average Recovery for Domain

Domain	Number of Runs Measured	Theoretical Length Being Measured (m)	Total Measured Core Length (m)	Average Recovery (%)
Breccia	383	727.43	549.2	76%
Fe-Oxide	826	1,539.31	993.6	65%
Acidified/ACID/SOAP	222	400.14	260	65%
Supergene	2,175	3,450.23	3,151.4	91%
Primary	4,239	4,657.60	4,600.7	99%
Primary Zn	1,903	2,688.06	2,653.92	99%
Total	9,748	13,462.77	12,208.82	91%

11.2.6

Geotechnical Logging

All diamond drill core from holes B-147 to B-309 was orientated using the Spear Method (a spear with a grease pencil attached to the tip is dropped down the hole to make a mark on the lower side of the core prior to pulling to core barrel). Each drill had two geologists/technicians assigned to the drills on a 24-hour basis to assure the quality of the orientation spear marks and to mark the core reference lines. The geologist/technicians also measured the core recovery and collected RQD measurements (noted in Section 11.2.5). No orientated core was completed prior to hole B-147.

Orientated core measurements were logged as interval data and point data using standardized codes for geotechnical determination of the rock parameters, including:

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-10
	October 2004

• Interval data.
  

• Basic data, drill run from and to.
  

• Rock fabric, weathering and strength.
  

• Rock mass, fracture frequency, structure set.
  

• Point Data.
  

• Depth, core circumference.
  

• Structure type.
  

• Orientation, alpha and beta angle of structure.
  

• Surface condition, planarity, roughness, infill and width of structure.
  

The orientated core measurements usually did not commence until a depth of 60 to 100 m downhole, after the core size was reduced to NQ diameter core and competent rock was encountered. No orientated core measurements were taken in the upper portions of the holes due to the inability to make suitable strike marks and/or piece the fractured core together.

The geotechnical data types and methodologies in use for capturing the data are in accordance with standard industry practices.

11.2.7

Dry Bulk Density Measurement

A total of 1,991 bulk density determinations (also referred to as specific gravity, “SG”, herein) on drill core have been completed by Nevsun and at ALS Chemex. Data was collected for all rock types, mineralization types and grade ranges. Nevsun further compiled these values into the six identifiable geological domains and average SG values were calculated within these domains (see Table 11-5).

Table 11-5
Nevsun Bulk Density Classification by Geological Domain for 2003

Geological Domain	Bulk Density (g/cm³ )
Iron Oxide (OXID)	3.19
Acid Rock (ACID)	2.18
Supergene Massive Sulphide (Supergene)	4.26
Primary Massive Sulphide (Primary)	4.42
Stringer Zone (STSX)	2.96
Meta-volcanics (BX)	2.61

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-11
	October 2004

Samples for each domain were collected from HQ and NQ diameter drill core. Nevsun sample collection and density value determination evolved over time with the use of improved procedures and better equipment from the first drill program in 2003 to the latest drill program 2004. Changes to procedures improved quality and precision of the values.

The bulk density determinations by Nevsun in Phase I 2003 drill program were made on air-dried, unsealed core using a beam microbalance and a standard immersion-displacement technique, from which the bulk densities were calculated. From the 260 samples collected at site, 44 of the same samples were also sent to ALS Chemex, Vancouver, Canada for an independent measurement and comparison of the results. The method in use by ALS Chemex was to oven dry the samples, coat in paraffin wax, weigh the sample in air and in water to determine the bulk density value. The comparison of the bulk density values resulted in procedural changes. Principally, the method used small pieces of core and was considered to be prone to over-estimation of the density values. Nevsun improved the quality of the measurements by using larger samples and purchasing more accurate equipment.

The bulk density measurements collected at site during the Phase II 2003 drill program were made on oven-dried, unsealed core using an electronic balance and standard water immersion displacement technique. A total of 611 samples were collected from site during this program, of which 68 were sent to ALS Chemex for independent measurements and subsequent comparison. The comparison continued to reveal problems with the porous rock types from the oxide domains, such as SOAP and FERU (see Section 7.0). Nevsun therefore made further changes to equipment and procedures for the Phase I 2004 drill program. The changes included the purchase of a bulk density scale and coating the samples with paraffin wax. During the 2004 program Nevsun measured 311 samples and submitted 697 samples to ALS Chemex, of which 35 were from the original 311 samples measured at site as an independent check. The measurements collected at site and at ALS Chemex compared well.

A total of 1,991 samples were evaluated by AMEC for use in the database for bulk density of the domains in the resource model. The SG values were loaded into GEMS® Software using the coordinates of the samples down the drill hole and the points generated were all captured within the 3D wireframe and back tagged to their respective geological domains. The total count of SG values in all of the modelled domains is 960 samples (Table 11-6).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-12
	October 2004

Table 11-6
Bulk Density Samples in Domain and Considered for 2004 Resource Estimate

Domain	Samples Captured by Domain (count)	Acceptable Samples by Domain for the Resource Model (count)	Bulk Density (g/cm³ )	Minimum Value	Maximum Value
Breccia	32	15	2.21	1.77	3.19
Oxide	100	36	3.32	1.89	4.05
Acid	17	6	2.08	1.10	2.82
Supergene	233	115	4.15	1.38	5.05
Primary Zn	305	260	4.39	2.00	5.39
Primary Cu	273	176	4.43	2.32	4.99
Total	960	608

AMEC reviewed the data by separating the measurements by method, location, program and rock type. Descriptive statistics and comparison plots were prepared for each dataset to determine the average values and variability of the data. The 2003 dataset of on-site measurements lacked precision and had high variability for each domain. Only the values from the independent measurements sent to ALS Chemex in 2003 were accepted for use, for a total of 112 samples. The investigation of the data from the 2004 drilling program showed good comparisons to checks at ALS Chemex and had lower variability within domains. All data from site and the ALS Chemex measurement for the 2004 program were accepted for use.

Further details are provided in Section17.7 regarding the bulk density values used for the resource estimate.

11.2.8

Results

Nevsun completed 310 diamond drill holes for 48,309.66 m on the property as of the completion of the June 2004 drill program (See Table 11-1 and Appendix C, Tables of Drill Hole Data). There are 471 significant drill hole intercepts (an intercept often included more than one of the geologic domains) of mineralization on the property (See Table C-2 in Appendix C, Tables of Significant Drill Hole Intercepts). The drilling has identified the mineralization at Bisha Main through to Bisha South and also in the Northwest Zone.

The mineralization on the Bisha Main and South Zones was intercepted by 253 diamond drill holes over a strike length of 1,200 m. These holes produced 683 mineralized intervals from the six geological domains (Table 11-7).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-13
	October 2004

Table 11-7
Mineralized Intervals for Each Geological Domain

Domain	Diamond Drill Hole Intercepts	Number of Mineralized Intervals
Breccia	61	63
Fe-Oxide	94	104
Acidified	87	91
Supergene	147	152
Primary	138	165
Primary Zn	114	121

Collar spacing on the Bisha Main and South zones are approximately 25 m apart along the strike from 1715100 North to 1715250 North and drilling has intercepted mineralization to a maximum depth of 380 m below the surface. The majority of the holes, especially in the delineation of the Main Zone were angle holes drilled from west to east for both the eastern and western synclines. A few holes were drilled east to west for better delineation of the western syncline. The azimuth of the angled holes was oriented parallel to the grid lines, and at dip angles designed to intersect the mineralization at various depths. A similar drill hole layout was used throughout the drilling of the Main Zone through to the South Zone. A small number of vertical holes were drilled over both zones.

11.3

RC Drilling

The reverse circulation drilling program was carried out during the 2004 Phase I program using a truck mounted, centre-sample return, triple-wall system UDR-650-P35 combination drill, owned and operated by Major Pontil Pty Ltd. of Australia.

A total 2,406.10 m (approximately 4% of drilling) was drilled in 42 holes during the program with a total of 2,097.3 m of RC and 308.8 m of core drilling (Table 11-1). All descriptions of the core handling and data are provided in Section 11.2.

11.3.1

Collar Surveys

Refer to Section 11.2.1 for a description of the collar survey practices and results.

11.3.2

Downhole Surveys

No downhole surveys were completed for azimuth measurements on the reverse circulation portion of the BRC drill holes. Reflex measurements of the azimuth and dip were completed near the end of cored intervals of four of the combination drill holes (BRCD-032,033,034,038).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-14
	October 2004

The Reflex unit was also used to collect 39 dip measurements within most of the BRC holes.

11.3.3

Logging

Samples were collected in 2 m intervals. The samples passed through a conventional cyclone and were collected in pails before they were passed through a riffle splitter (two stage SP-2 Porta Splitter). Approximately 10% of the original sample (2 kg) was obtained after the riffle splitting (2 to 4 passes depending on the original volume of material recovered).

Representative chips were collected from the reject material and were logged on site with about 25 to 50 g of the sample archived in plastic chip trays. All key aspects of the geological data were captured on the logs, including:

• Major lithology
  

• Subsidiary lithology
  

• Alteration
  

• Mineralization.
  

The remainder of the sample material was discarded. The chip trays were labelled and stored in a locked storage container located at the Nevsun camp.

11.3.4

Recoveries

The sample volume, weight and number of splits were recorded for each sample, both at the drill (wet sample) and later in camp (dry sample) in an effort to determine sample loss.

Significant discrepancies may occur for the recovery of RC material due to the variations in several parameters related to the drilling method. The calculation used to determine the maximum theoretical weight for each 2 m sample is based on a supposedly known volume of rock. The rock type, and therefore bulk density, can vary within a 2 m sample length. Thus, sample weights could actually be reported to be larger or smaller than the theoretical 100% weight. Nonetheless, recoveries were poor for all drill holes. The values ranged from 0 to 61% and resulted in an average recovery of 16% for all the RC samples.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-15
	October 2004

Samples intervals from RC drilling comprised only approximately 4% of the drill hole database and of this only 3.5% of the significant mineralized intervals in the geologic model are from RC drilling (Table 11-8). Although AMEC did not exclude these holes from the resource estimate, the relative impact of the holes will be minor and the average grades of these intervals are lower than core holes. The relative impact of the use of RC holes should be further assessed during any further resource estimates.

Table 11-8
Distribution of Sample Intervals by Domain for each Hole Type

	Metres from Drill Hole Type in Resource Model				Percent Mineralization from Drill Hole Type in Resource Model
Domain	B	BRCD	BRC	Total	B	BRCD	BRC
Breccia	678.00	-	141.00	819.00	82.8%	0.0%	17.2%
Fe Oxide	1,476.89	26.63	177.40	1,680.92	87.9%	1.6%	10.6%
Acidified	379.41	13.80	27.00	420.21	90.3%	3.3%	6.4%
Supergene	3,402.52	30.36	94.50	3,527.38	96.5%	0.9%	2.7%
Primary	3,357.15	-	3.00	3,360.15	99.9%	0.0%	0.1%
Primary Zn	2,750.66	-	-	2,750.66	100.0%	0.0%	0.0%
Total	12,044.63	70.79	442.90	12,558.32	95.9%	0.6%	3.5%

11.3.5

Results

Nevsun completed 42 reverse circulation drill holes for 2,097.30 m (including 9 combination core/RC holes) on the Property as part of the June 2004 drill program (See Table C-2 in Appendix C). The RC drilling was confined to resource definition drilling on the Bisha Main and Bisha South Zones. The RC drilling resulted in 41 mineralized intervals from the six geological domains (Table 11-9).

Table 11-9
Mineralized Intervals for Each Geological Domain

Domain	Reverse Circulation Hole Intercepts	Number of Mineralized Intervals
Breccia	10	10
Fe-Oxide	17	18
Acidified	4	4
Supergene	8	8
Primary	1	1
Primary_Zn	-	-

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-16
	October 2004

Most of the drill holes completed by the RC drill were located along the periphery of the deposit, targeting lower priority areas and many were shallow drill holes.

11.4

Water Well Drilling

Twenty-one holes for water wells and to provide information on ground water levels and flow data have been drilled on and nearby the Bisha Property. The data is not complete for all the water wells (Table 11-10); therefore, the total metres drilled for water wells are unknown and the exact location of Well #6 is not documented.

Table 11-10
Summary of Water Well Locations

Hole Id	Northing	Easting	Elevation	Depth (m)	Drill Date	Purpose	Company
Well 1	1719807.00	337850.00	540.00	50	2003	Production	Eritrean Drilling
Well 2	1715883.00	339183.00	566.00	40	2003	Abandoned	Eritrean Drilling
Well 3	1716097.00	339221.00	565.00	50	2003	Abandoned	Eritrean Drilling
Well 4	1716134.00	339201.00	565.00		2003	Abandoned	Eritrean Drilling
Well 5	1715884.00	339197.00	563.00	33	2003	Abandoned	Eritrean Drilling
Well 6					2003	Production	Eritrean Drilling
BHMW04-01	1715679.04	339539.07	563.24	60.8	2004	Monitoring	Major/Eritrean
BHMW04-02	1715922.94	339528.93	562.36	82	2004	Monitoring	Major/Eritrean
BHMW04-03A	1715288.34	339324.23	565.29	80	2004	Monitoring	Major/Eritrean
BHMW04-03B	1715288.34	339324.23	565.29	20	2004	Monitoring	Major/Eritrean
BHMW04-04A	1715648.87	339222.69	562.70	79.7	2004	Monitoring	Major/Eritrean
BHMW04-04B	1715648.87	339222.69	562.70	20	2004	Monitoring	Major/Eritrean
BHMW04-05	1716099.73	339150.20	559.96	50.8	2004	Monitoring	Major/Eritrean
BHMW04-06A	1716285.16	339272.82	563.74	59.6	2004	Monitoring	Major/Eritrean
BHMW04-06B	1716285.16	339272.82	563.74	20	2004	Monitoring	Major/Eritrean
BHMW04-07	1716153.28	339427.69	570.55	17	2004	Monitoring	Major/Eritrean
BHW05	1714004.74	340473.33	574.17	70	2004	Production	Eritrean Drilling
BHW06	1713287.85	342488.34	581.22	57	2004	Production	Eritrean Drilling
BHW07	1718976.77	337536.63	545.95	82	2004	Production	Eritrean Drilling
BHW08	1713168.05	340063.79	575.04	33	2004	Production	Eritrean Drilling
BHW09	1717947.83	336483.38	549.54	36	2004	Production	Eritrean Drilling

Six water well holes were completed in 2003 (Well 1 to 6). Two of these holes are used to provide water. One well supplies water for camp use and the other well is used for drill operations. Eritrean Core and Water Well Drilling, a local Asmara contractor, completed the drilling. The equipment included an Atlas Copco Aquadrill R5C and separate XRHS 385 compressor with a working pressure of 16 bars. Each hole was drilled with an 8” hammer bit and lined with 6” plastic perforated pipe. The space between bore and casing was filled with -0.5 cm size screened gravel (Nevsun, 2004).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-17
	October 2004

The water well drilling completed during the 2004 program was to assist in the capture of ground water data near the Bisha mineralization and to provide data on water production for potential mining and processing activities. Fifteen holes for total of 767.9 m were completed by the local Asmara based contractor, Eritrean Core and Water Well Drilling, as well as by Major Pontil Pty. Ltd. The drilling of the wells was under the supervision of Klohn Crippen Consultants Ltd. who also trained personnel to monitor water levels (Nevsun, 2004).

Seven of the monitoring wells (BHMW04-01 to BHMW04-07) were drilled on the periphery of the proposed open pit area. An upper and lower piezometer was installed in each monitoring well. At three of the well sites, a second shallow hole adjacent to the original hole, was drilled to install the upper piezometer.

The five other holes were drilled to assess potential for production purposes (BHW05 to BHW09). Not all of these holes are within the concession boundaries. One water well, located several kilometres to the northwest of the Bisha Main deposit and on the northern concession boundary, intersected a flow rate of 15 L/s. Additional wells with high flow rates are also located close to the Barka River some 15 km to the north of the property.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 11-18
	October 2004

12.0

SAMPLING METHOD AND APPROACH

12.1

Introduction

Sampling programs at the Bisha Property included drill core samples, RC samples and arious geochemical samples, which included: surface rock chip, trench, auger, pit, soil and stream sediment sampling.

Nevsun established detailed logging, sample collection and sample preparation protocols for core and RC sampling, and implemented procedures for the collection of geotechnical data. Documentation of geochemical sampling procedures are in most cases incomplete, except for Mercier’s report (2003) entitled “Geochemical Surveys, a Final Report Bisha and Okreb Prospecting Licenses Area” that documents the procedures adopted for the collection of samples during the 2003 Phase I exploration program.

AMEC was on site during core and RC drilling and was able to observe the procedures for those activities.

12.2

Soil Sampling Procedures

12.2.1

Nevsun Soil Sampling Procedures

A total of 6,287 soil samples have been collected to date on the Bisha property (Figure 10-4). The sampling focused on several target areas on the Bisha property, including:

• Bisha Main Zone.
  

• Bisha South (now included as part of the southern portion of the Bisha Main Zone).
  

• SE Anomaly.
  

• NW Barite Anomaly.
  

• Harena Anomaly.
  

Soil samples were first collected over the Bisha Main Zone mineralization (Table 12-1) in 1999, and again in 2003 with a significant number of samples collected in 2004 over the entire Bisha Main and South areas. Much of the sampling was follow up to the anomalies defined by geophysical surveys and to build on the exploration dataset.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-1
	October 2004

Table 12-1
Total Line Kilometres of Soil Sampling on Bisha

Target	Pre 2004	Coverage (line km) 2004	Total
Bisha Main	28.42	74.3	102.72
Barite	5.12	22.32	27.44
Harena	3	17	20
Exploration	1.78		1.78
Total (line Km)			151.94

The 1999 soil sampling was completed over three gridlines on a grid established using an azimuth of 010° (from magnetic north) with gridlines spaced 200 m apart. The spacing of the samples along the section lines is not known to AMEC. The samples were sieved to -80 mesh. There was no documentation available to AMEC regarding from which horizon the samples were collected or the size of sample.

Sampling in 2003 and 2004 was completed on grid lines over the Bisha Main, Harena and NW Barite Showing areas. The gridlines were established using a Trimble Pro-Mark 2 GPS system consisting of a base station and rover unit with a radio link. The GPS unit is capable of sub-meter accuracy.

Samples from the Bisha Main Zone were collected from gridlines spaced 50 to 100 m apart. The samples were collected at 25 m intervals along the gridlines. Sampling at the Harena Area was on gridlines spaced 200 m apart with two 400 m gaps on a grid baseline that trends 035°. The samples were collected at 25 m stations along the lines. The NW Barite Showing area had sampling gridlines spaced 200 metres apart and sample stations were at 25 m intervals along east-west oriented gridlines. An additional line of sampling was completed 7 km east of the Bisha Main Zone (at the north end of the Bisha Gabbroic Complex) at 25 m spacing along the line (Figure 10-1).

All of these samples were taken approximately 10 cm below surface regardless of the material type at the target depth. The samples were screened using a –60 mesh and 100 to 200 g of sample was placed in a pulp or kraft sample bag and labelled with the grid coordinates. Sample descriptions included separation of the soils into three distinct populations based upon colour and grain size (Nevsun, 2004). Red soils were classified as residual, while brown-coloured soils were classified as transported. Soil samples consisting of predominantly sand-sized grains were classified as alluvial. After returning the samples to camp, sample tags were attached to the sample bag prior to shipment to the laboratory.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-2
	October 2004

12.2.2

Mercier Soil and Auger

A total of 39 samples (23 soil samples and 16 auger samples) were collected in order to test the hand auger as a geochemical sampling tool. To do so, it was decided to locate with a GPS the previous sampling sites on line 1716150 of the 1999 campaign. Soil samples were collected at the previous locations. Furthermore, where it was possible (for obvious reasons, it was not possible to use the auger on the gossan area and where outcrops were present), it was decided to twin the soil sample with an auger sample (16 auger samples were collected; Mercier, 2003). For this purpose, a Dutch Auger unit was purchased (with 6 m of extension rods).

Approximately 2 to 3 kg of sample material was collected at the bottom of the auger holes.

The auger sampling was not determined to be advantageous and therefore was not continued.

12.2.3

Termite Mound Sampling

Termite mounds are present in the Bisha Property. The termites transport residual soil from depth to the surface of their mounds.

A total of 107 termite mound samples (including four duplicates) and 8 auger samples of the mounds were collected in the area of the Bisha Prospecting License. The area sampled covers approximately 55 km² (Mercier, 2003).

One team was used for the sampling: a geologist was in charge of three unskilled workers. To detect the mounds, traverses were planned in a north-south direction 500 m apart. The samples were collected in the upper part of the mound (the more recent material deposited). Approximately 8 kg of sample was then put in a bag. A field coding form was used to record sample number, UTM coordinates, site description, etc.

The termite sampling provided some additional geochemical information but were limited to areas with mounds.

12.2.4

Pit Sampling

Twelve pits were excavated with a backhoe along a 7 km long line to make a short investigation of the different regolith units. Three samples were collected in each pit, with the exception of pits 406, 407 and 412, where only two samples were collected. The first sample (sample A) was collected in the upper part of the profile and was an equivalent of a soil sample collected at the surface; the second sample (sample B) was collected in a different horizon than the one observed in the upper part of the pit and corresponded normally to the transition zone between the surface material and the bedrock; the third

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-3
	October 2004

sample (sample C) was collected among fragments of bedrock or what was generally recognized to be the beginning of the oxidized bedrock. A total of 33 samples were collected in the 12 pits (Mercier, 2003).

12.3

Rock Chip Sampling Procedures

A total of 461 of rock chip samples were collected during mapping and prospecting of the Bisha Property. The sample database lacks documentation of the type of sample for most of the samples (i.e. whether the samples were channel chip, grab or float samples). UTM coordinates were collected for all sample locations using a hand held GPS instrument. The rock type, alteration and m ineralization was noted at the sample location for all the sample types.

12.4

pH Survey Procedures

During the 2003 Phase II exploration program the use of the pH measurement of soils was investigated to determine whether it would be a useful method to define sulphide mineralization. The sampling was carried out on seven gridlines: 1715200N, 1715550N, 1715650N, 1716000N, 1716100N, 1716600N and 1717000N. Samples were collected at 25 meter spacing between 339325E and 339475E for lines 1715550N and 1715650N. The samples on the other lines were at 25 m spacing at various locations on the gridlines (Chisholm et. al., 2003). The total numbers of samples or line kilometres was not documented.

Initially two gridlines acted as orientation lines to determine the relative merits of this type of survey. Two sets of samples were collected from the same sample location, one near surface at less than 10 cm depth and the other at approximately 25 cm depth. After evaluation of the results, Nevsun completed the remaining sampling on the other five lines at a depth of 10 cm.

The samples were collected using a hoe or shovel tool and sieved with a –28 mesh. Approximately 100 g to 200 g of sample was collected in a kraft sample bag and labelled with the appropriate coordinates. The samples were returned to the Nevsun camp for measurements on the same day of collection.

12.5

Stream Sediment Sampling Procedures

Nevsun carried out a stream sediment sampling program in 1998 and again during the 2003 Phase I exploration program. The 2003 work was designed and implemented under the direction of M. Mercier, an independent geochemical consultant. It should be noted that all streams on the Bisha Property are seasonal.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-4
	October 2004

The 1998 survey included collection of 127 samples from category 1 to category 4 streams located on the Bisha Property. The samples were sieved using a –28 mesh and an unknown quantity of sample was shipped for analyses. Sample locations were recorded using a hand held GPS unit.

A stream sediment survey with a total of 165 samples was completed over an area of 165 km² in 2003. In the Bisha area, the samples were collected in category 1 and 2 streams. Occasionally, streams of higher order (3 and 4) were also sampled for a comparison study (Mercier, 2003).

One team was used for the sampling: a geologist was in charge of three unskilled workers. The samples were collected in pits across the active bed of the stream and sieved at the sampling site at minus 1mm. Approximately 25 kg of composite sample was then put in a rice bag. A field coding form was used to record sample number, UTM coordinates, site description, sample depth, etc. (Mercier, 2003).

12.6

Trench Sampling Procedures

A series of 36 trenches were excavated with a Komatsu WB93R utility tractor equipped with a backhoe over various parts of the Bisha Main and Northwest Zones (no figures are provided for the trenches in this report). The trench samples were not included in the database for resource estimation.

The trenches were excavated to a depth of 0.5 m to 3.5 m depending on the difficulty of excavation or breaking the rock. A total of 707 samples were collected from 1,402 m of excavated trenches as summarized in Table 12-2.

The trenches were mapped and then sampled by the same geologist. Channel samples were taken at 2.0 m intervals and respected lithologic contacts. The rock type as well as alteration and mineralization from the sample location was noted for all the sample types.

12.7

Drill Core Sampling Procedures

The Bisha Property was drilled using three types of diamond drill rigs having varied core barrel diameters. The core sizes from smallest to largest diameter were: BTW (41.0 mm), NQ (47.6 mm), NTW (55.1mm) and HQ (63.5 mm). Both NQ and HQ cores were produced by the Universal Drill Rig (RC and diamond drill combination unit; pers. comm. Nielsen, 2004).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-5
	October 2004

Table 12-2
Summary of Trench Locations

Trench	Easting	Northing	Elevation	Azimuth	Inclination	Length (m)	Samples Collected
BT-01	339439.20	1716070.04	571.01	112.0	0.0	7	6
BT-02	339392.11	1715908.36	563.79	107.0	0.0	30	17
BT-03	339296.43	1716040.74	565.13	97.0	0.0	12	10
BT-04	339405.01	1715931.13	566.18	102.0	0.0	31	17
BT-05	339296.24	1716022.40	564.28	104.0	0.0	24	11
BT-06	339436.14	1716081.90	572.11	94.0	0.0	17	8
BT-07	339284.74	1716127.96	569.43	121.0	0.0	41	21
BT-08	339107.68	1716091.63	562.84	9.0	0.0	23	2
BT-09	339095.73	1716090.20	562.75	12.0	0.0	23	3
BT-10	339407.75	1715987.29	563.49	104.0	0.0	126	66
BT-11	339386.37	1715806.41	564.23	103.0	0.0	57	29
BT-12	339311.31	1716066.17	567.30	91.0	0.0	19	8
BT-13	339110.05	1716085.80	563.75	100.0	0.0	13	6
BT-14	339127.54	1716097.76	562.54	169.0	0.0	7	0
BT-15	339143.22	1716073.42	563.80	29.0	0.0	6	2
BT-16	339129.68	1716080.97	563.01	29.0	0.0	9	4
BT-17	338320.46	1717479.30	554.85	83.0	0.0	42	21
BT-18	338285.90	1717512.35	553.85	55.0	0.0	40	21
BT-19	338267.35	1717546.28	554.37	63.0	0.0	23	12
BT-20	338232.26	1717578.46	553.90	70.0	0.0	24	12
BT-21	338337.91	1717506.91	555.65	110.0	0.0	173	91
BT-22	338556.92	1717778.80	551.37	128.0	0.0	80	41
BT-23	339145.31	1715445.33	558.81	102.5	0.0	48	23
BT-24	339177.19	1715496.75	559.06	104.5	0.0	41	18
BT-25	339331.32	1715712.19	558.00	125.5	0.0	15	8
BT-26	339314.26	1715728.26	557.77	5.0	0.0	23	12
BT-27	339349.54	1715715.98	557.74	43.0	0.0	24	12
BT-28	338675.43	1717797.87	553.03	90.0	0.0	25	13
BT-29	339514.69	1716367.17	645.26	113.3	0.0	5	3
BT-30	339513.38	1716363.23	643.70	108.3	0.0	3	2
BT-31	339415.00	1715238.00	562.00	90.0	0.0	36	18
BT-32	339187.55	1716168.83	558.48	87.1	0.0	31	19
BT-33	339215.69	1716102.53	557.53	90.0	0.0	228	119
BT-34	339163.07	1716034.29	560.95	80.6	0.0	37	21
BT-35	339146.60	1716074.34	560.29	32.6	0.0	13	8
BT-36	339301.58	1714800.97	564.54	90.0	0.0	46	23

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-6
	October 2004

The Kluane drill rig (details in Section 11.2) collared holes with NTW and reduced to BTW, while the Boart Longyear drills began with HQ diameter core and reduced to NQ. Not all holes were reduced if the ground conditions permitted reasonable penetration or if ground conditions (badly fractured) were not favourable for reduction to a smaller core diameter. The drillers made the decision when to reduce, under direction and approval from the Nevsun site managers. The drilling types and drill diameters used are summarized in Table 12-3.

Table 12-3
Summary of Drill Techniques Used

Drill Type/Diameter	Drill Hole Count	Metres Drilled	Percentage
NTW	17	1,641.72	3.4%
BTW	13	768.01	1.6%
HQ	292	16301.95	33.5%
NQ	269	29597.98	60.9%
BRCD	9	308.70	0.6%

All of the smaller diameter drilling reflects a reduction in drilling diameter in a drill hole collared as NTW or HQ.

The first 17 holes, B-01 to B-16 and B-2a did not have a geologist/technician at the drill site during the drilling; otherwise the sampling methods are the same as the later holes (B-17 to B-309) discussed below. Recovery data and Rock Quality Designator data (RQD) was not collected until hole B-13. The Nevsun drill core handling procedures for drill holes B-17 to B-309 are summarized below.

• Prior to transport, all diamond drill core is reviewed for any run block irregularities and measured for core recoveries and RQD determination by two geologists/technicians assigned to the drills on a 24-hour basis. During the 2004 drilling, the geologists/technicians checked the quality of the orientation spear marks and mark the core reference lines.
  

• After the core arrives in camp, it is washed and meterage blocks are again checked by the geologist to ensure no error is present in the runs from the drill.
  

• The geologists at the core area complete geological and geotechnical logging on hardcopy log forms using a standard library of lithological, alteration, mineralization codes. The geologist determines the sampling interval and this is identified by ink marks on the core and with paper tags being placed under the core within the interval.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-7
	October 2004

• The maximum sample length is 12.00 m (only within wallrock away from mineralized intervals) and the minimum is 0.15 m. Within the zones of mineralization, samples lengths are generally between 1.00 and 3.00 m. Sample intervals are determined based upon mineralogical and lithologic contacts (Table 12-4).
  

• The core is laid out for digital photography and then removed to the core storage area if no samples are marked or if samples are marked, the core is sent to the core cutting area.
  

Table 12-4
Summary Statistics of Sample Lengths Grouped by Rock Type

		Lengths in Metres
Rock Code	N Samples	Minimum	Maximum	Mean
ALUV	64	0.30	3.00	1.52
BRX	17	1.00	1.50	1.43
CHRT	5	0.50	1.50	0.81
CONG	47	0.20	12.00	1.74
FBPD	26	0.60	1.60	1.31
FELD	241	0.20	3.20	1.41
FELT	542	0.30	8.00	1.62
FELU	53	0.20	3.00	1.65
FERC	379	0.20	6.00	1.61
FERU	390	0.20	3.25	1.53
FLT	7	1.50	1.50	1.50
FPDK	145	0.35	3.00	1.42
FTBX	11	0.50	1.50	1.40
GOUG	1	1.60	1.60	1.60
HALF	239	0.40	3.20	1.51
HEBX	126	0.50	3.00	1.65
INTD	11	1.00	2.03	1.43
INTF	7	0.70	1.50	1.19
INTT	77	0.80	2.00	1.37
MAFD	99	0.25	10.50	1.32
MAFF	108	0.75	2.00	1.43
MAFT	3169	0.40	6.00	1.51
MAFU	3	1.35	1.50	1.45
MDST	124	0.30	3.00	1.30
MSUL	6310	0.20	4.30	1.46
NR	50	1.50	9.00	2.04
OVBD	10	1.50	1.50	1.50
QFPD	29	0.60	1.50	1.28
QPDK	50	0.70	3.00	1.57
QZBX	283	0.20	6.00	1.68
REBX	65	3.00	3.00	3.00
SAND	11	1.30	2.80	1.62
SAPR	2027	0.40	6.00	1.54
SHR	2	1.50	1.50	1.50
SMSX	233	0.40	3.15	1.31
SOAP	2468	0.15	9.00	1.54
STSX	2797	0.30	3.80	1.45
VNQZ	26	0.20	3.90	1.57
VOID	4	2.95	3.00	2.98
XTUF	108	0.50	1.70	1.38

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-8
	October 2004

• Standard diamond cutting blades flushed with water are used to half the core. Highly broken core pieces are cut along the axis if possible or the core is split using a trowel down the middle of the tray row and hand picked or scooped to ensure representative samples are obtained. Cutting lines are not drawn on the core because of the massive nature of the mineralization. Generally the mineralization is lacking any significant banding or veining. The remaining half core is returned to the core storage area and stacked in the numerical order of the core box numbers.
  

• The core splitters place half of the core in double lined plastic bags with the sample tag placed inside of the bag and the sample number labelled on the outside of the bag.
  

• The open bags are placed outside of the lab in a secure area on a gravel pad to dry in the sun. For holes B-01 to B-53, samples were sealed and packed in hard plastic blue pails. Each pail contained 15 to 20 samples. The pails were securely sealed by drilling holes in the lid of the pail and ‘zap-strapping’ the lid to the pail. For samples from holes B-54 onwards, the preparation laboratory operated by Nevsun took control of sample preparation prior to shipping of the samples (see Section 13.0).
  

AMEC observed the current drilling, geotechnical, geological, core sampling and sample preparation on site and considered that the practices employed are acceptable and according the standard industry practices.

12.8

Reverse Circulation Drill Sampling Procedures

The reverse circulation drill program was carried out during the 2004 program using a truck mounted, 136 mm diameter, centre-sample return, triple-wall system UDR-650-P35 combination drill, owned and operated by Major Pontil Pty Ltd. of Australia (Nevsun, 2004).

The procedures as described within the Nevsun 2004 Program report are as follows:

• Each sample interval was 2 m in length.
  

• While drilling, the sample that passed through a conventional cyclone, was collected in pails and then passed through a riffle splitter (two stage SP-2 Porta Splitter).
  

• Approximately 10% of the original sample (2 kg) was obtained after the riffle splitting. The remaining sample was discarded.
  

• At the end of each drill shift the samples were transported to Nevsun’s camp and deposited at the sample laboratory on the gravel pad.
  

• The samples were sorted and dried in the sun in a secure area before placing in the prep lab oven.
  

• Nevsun preparation laboratory took control of the samples prior to shipping.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-9
	October 2004

AMEC observed the RC drilling and sampling while on site. The practices employed are acceptable and according the standard industry practices. AMEC also recommends that the reject material should be retained for future reference. Also, several additional parameters should be recorded such as depth to water table, number of splits of the original sample, etc.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 12-10
	October 2004

13.0

SAMPLE PREPARATION, ANALYSES AND SECURITY

13.1

Introduction

Exploration work during 1998 and 1999 used the Intertek Testing Services Bondar Clegg Laboratory (ITS Bondar Clegg), based in Asmara. There is little documentation provided in the reports reviewed by AMEC that documents the sample preparation and the analytical protocols from the earlier work.

All trench, rock chip and geochemical samples, including soil and auger, stream sediment, pit and termite mound samples collected during the 2003 Phase I program were shipped to the Horn of Africa Preparation Laboratory, in Asmara, which provided provides preparation services for Genalysis Laboratory Services Pty (Genalysis) of Maddington, Australia. The preparation laboratory produced pulp samples that were subsequently shipped to Genalysis in Australia for analysis. Following the 2003 Phase I program, geochemical and rock chip samples were shipped to ALS Chemex Ltd. (ALS Chemex), in Vancouver, Canada.

The primary laboratory used by Nevsun for analytical work on the drilling programs was ALS Chemex. Nevsun used the laboratory for both sample preparation and analyses since the initiation of the first drill program in 2002. During the 2002 and Phase I of the 2003 drilling program samples were shipped as half-core from the Bisha camp to Asmara and forwarded to ALS Chemex in Vancouver via Lufthansa Airlines. After establishing a sample preparation facility¹⁹ at camp in September 2003, Nevsun sent coarse crushed and split material (-2 mm) for core, RC, and rock samples to ALS Chemex for subsequent pulverization and analyses. All assay data contained in the database for resource estimation was assayed by ALS Chemex.

Both ALS Chemex and Genalysis are ISO registered and are internationally recognized facilities. ALS Chemex is registered to ISO 9001:2000 for the “provision of assay and geochemical analytical services” by BSI Quality Registrars (Appendix D, ALS Chemex Analytical Procedures). The National Association of Testing Authorities Australia has accredited Genalysis, following demonstration of its technical competence, to operate in accordance with ISO/IEC 17025 (1999), which includes the management requirements of ISO 9002:1994. The facility is accredited in the field of Chemical Testing for the tests, calibrations and measurements that are shown in the Scope of Accreditation issued by NATA (see Genalysis Website, 2004).

The ITS Bondar Clegg Laboratory, based in Asmara is no longer in existence. Internationally, ALS Chemex took over Bondar Clegg in December 2001.

¹⁹  The sample preparation facility was designed and assembled by ALS Chemex for Nevsun.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-1
	October 2004

13.2

Sample Preparation for Soils and Sediment

Nevsun has utilized three separate labs (ITS Bondar Clegg, Genalysis and ALS Chemex) for geochemical work since the beginning of exploration on the Bisha Property. ALS Chemex has completed the largest amount of preparation and analytical work.

The stream sediment samples from 1998 and soil samples collected in 1999 were sieved on site by Nevsun with –28 and –80 mesh sieves, respectively. The samples were shipped to the ITS Bondar Clegg Laboratory, based in Asmara for processing (pers. comm. Bill Nielsen, 2004).

13.2.1

Nevsun – ALS Chemex

In 2002, Nevsun used a –60 mesh to screen the samples in the field and the fine fraction was retained for analysis. This is considered satisfactory for smaller (i.e. 500 g or less) samples where the exploration target is base metals (Mercier, 2003). ALS Chemex did no further preparation of the sample prior to digestion and analysis.

Usually when gold is the exploration target, the particle size of the fine fraction should be further reduced using ring mill pulverization, i.e. to > 85% - 75 microns (150 mesh) in order to obtain more reproducible gold data (ALS Chemex website, 2004; www.alschemex.com). However, gold analytical results provided poor results on the Bisha Main even though the gossan has gold values. Nevsun maintained the –60 mesh screen size.

13.2.2

Mercier – Horn of Africa

A geochemical program designed, implemented and supervised by M. Mercier during the 2003 Phase I exploration program used the Horn of Africa Preparation Laboratory to prepare samples. Mercier provided instruction on the preparation methods for all the samples and observed the preparation of the first samples in the lab. The remaining samples were prepared under the supervision of a Nevsun geologist. All samples were shipped to Genalysis, Australia for analysis.

13.2.3

Stream Sediment Sample Preparation

The instructions by Mercier for stream sediment sample preparation were as follows (Mercier, 2003):

• A riffle splitter was used to quarter the samples.
  

• Approximately one quarter (about 7 kg to 8 kg) of the sample was sieved (a bigger quantity was taken for the samples to be analyzed for the Platinum Group Elements or if there was not enough material after quartering).
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-2
	October 2004

• Sieved at minus 80 mesh (-180 µm).
  
• All samples were pulverized (the specification for final pulverization was >90% of the sample must be less than 200 mesh or 75 µm).
  
• Fraction +80 mesh was discarded.
  
• Fraction -80 mesh
  

-

150 g prepared for the analysis of Au and ICP.

-

150 g prepared for the analysis of Platinum Group Elements (samples B-ST-153, B-ST-159, B-ST-164, B-ST-174, B-ST-176, B-ST-184, B-ST-185, B-ST-190, B-ST-191, B-ST-192, B-ST-195, B-ST-200, B-ST-202, B-ST-203, B-ST-206, B-ST-207, B-ST-208, B-ST-214, B-ST-215, B-ST-216, B-ST-219, B-ST-221, B-ST-223, B-ST-224 and B-ST-228).

-

Remainder of the pulp was placed in a plastic bag identified with a label and an aluminum tag, and sent back to Nevsun.

13.2.4

Soil and Auger Sample Preparation

The instructions by Mercier for soil and auger sample preparation were as follows (Mercier, 2003):

• Entire sample was sieved.
  
• Sieved at minus 80 mesh (-180 µm).
  
• Fraction +80 mesh was discarded.
  
• Fraction -80 mesh
  

-

150 g prepared for the analysis of Au and ICP

-

Remainder of the pulp was placed in a plastic bag identified with a label and an aluminum tag, and sent back to Nevsun.

13.2.5

Pit Sample Preparation

The instructions by Mercier for preparation of pit samples ending with the letter “A” or “B” (see Section 12.2.2) were as follows (Mercier, 2003):

• Entire sample was sieved.
  

• Sieved at minus 80 mesh (-180 µm).
  

• Fraction +80 mesh was discarded.
  

• Fraction -80 mesh
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-3
	October 2004

-

150 g prepared for the analysis of Au and ICP.

-

Remainder of the pulp was placed in a plastic bag identified with a label and an aluminum tag, and sent back to Nevsun.

The instructions by Mercier for preparation of pit samples ending with the letter “C” (see Section 12.2.2) were as follows (Mercier, 2003):

• Crush the sample (>75% of the sample must pass 10 mesh or 2 mm screen).
  

• Splitting: only a portion of the crushed material was carried through to pulverizing stage. A crushed split of approximately 2 kg was derived from the crushing process using a riffle splitter. The rest of the sample was discarded.
  

• Pulverizing: 2 kg split was pulverized (the specification for final pulverizing is that >90% of the sample must be less than 200 µm).
  

-

150 g prepared for the analysis of Au and ICP.

-

Remainder of the pulp was placed in a plastic bag identified with a label and an aluminum tag, and sent back to Nevsun.

13.2.6

Termite Mound Sample Preparation

The instructions by Mercier for termite mound sample preparation were as follows (Mercier, 2004):

• Entire sample was sieved.
  
• Sieved at minus 80 mesh (-180 µm).
  
• Fraction +80 mesh was discarded.
  
• Fraction -80 mesh
  

-

150 g prepared for the analysis of Au and ICP.

-

Remainder of the pulp was placed in a plastic bag identified with a label and an aluminum tag, and sent back to Nevsun.

13.3

Sample Preparation of Drill Core and Rocks

13.3.1

Horn of Africa Preparation Laboratory

Rock chip and trench samples processed at the Horn of Africa Preparation Laboratory followed the following procedures:

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-4
	October 2004

• Samples sorted and ordered numerically after receipt.
  

• Placed in a drying oven for 12 to 18 hours at between 80°C and 100°C.
  

• Samples passed through a jaw crusher to >75% of the sample passing 10 mesh or 2 mm screen.
  

• Sample split using a riffle style splitter to a sub-sample size of between 200 to 250 g.
  

• Sub-sample pulverized with ring and puck pulverizer to >85% of the sample passing
  75 µm.
  

• Samples were shipped to Genalysis in Australia.
  

13.3.2

ALS Chemex

Rock and core samples sent to ALS Chemex prior to the implementation of the on-site sample preparation facility were prepared in the Vancouver preparation facility.

The samples were dried at 110-120°C for 10 to 12 hours and then crushed with either an oscillating jaw crusher or a roll crusher. The ALS Chemex quality control specifications for crushed material require that >70% of the sample must pass a 2 mm (10 mesh) screen (Figure 13-1).

The entire sample was crushed and typically 250 g was subdivided from the main sample by using a riffle splitter and carried through to the pulverizing stage. Generally ALS Chemex retains a 1-2 kg split of the reject in storage.

The 250 g split is pulverized using a ring mill. The ALS Chemex quality control specifications require that final pulverizing is >85% of the sample must pass 75 microns (200 mesh) (Figure 13-2).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-5
	October 2004

Figure 13-1
Particle Size Distribution Graph for Crushing

Figure 13-2
Particle Size Distribution Graph for Pulverizing

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-6
	October 2004

13.3.3

Nevsun Sample Preparation Laboratory

Nevsun purchased a fully equipped, containerized sample preparation laboratory from ALS Chemex in July 2003. The preparation laboratory was shipped from Vancouver and arrived on site in early October 2003. The lab was operational after minor alterations to the electrical wiring to connect to the generator (Photo A-27 in Appendix A).

The lab is controlled, operated and monitored by Nevsun staff and workers. The personnel working in the lab are mostly university educated and have prior experience working in laboratory conditions.

In late October, 2003, Gordon Walker from ALS Chemex, Turkey branch visited the lab to assist and review the start up and training of the personnel in sample preparation (crushing and splitting), equipment maintenance, QAQC, logistics and lab communication procedures (Appendix D).

Laboratory Type and Equipment

The laboratory is a 20 foot container (Photos A-26 and A-27 in Appendix A) outfitted with work benches, a drying oven (~6m³), two large, wheeled sample racks for the drying oven, an air compressor with hoses and air guns, one used and one new T.M. Engineering Rhino jaw crushers, a spare parts kit, a 2 mm sieve for crusher quality control, one riffle splitter and pans. AMEC reviewed all equipment and found it to be in good to excellent condition.

Sample Procedures and Processing

Samples are received directly from the core cutting area in plastic sample bags. The sample bags are laid out in numerical order with the plastic bags open at the top to aid in drying. The sample sorting area is a covered pad of crushed rock (Photo A-28). The weather is hot and sunny during the day.

Any hygroscopic saprolite samples were placed into sample pans so that they would dry more quickly. At the end of the day the samples are loaded into the drying oven to dry overnight at a temperature of 80°C (Photo A-26). The total drying time is between 11 and 12 hours.

Sample weighing is carried out first thing in the morning as the samples were taken out of the drying oven and had sufficient time to cool. The samples are laid out on the crushed rock pad in a secure area to be readily accessible for crushing (Photo A-26). The dry weights are recorded. The half core samples weigh between 2 kg and 11 kg. The half-core samples were usually double (most samples) or triple crushed (harder samples) using the

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-7
	October 2004

T.M. Rhino Jaw crushers to achieve the crushing target of greater than 70% passing –2 mm.

After crushing the first sample of the shift it was screened to ensure that the crushing protocol was being met. Other, random samples were selected throughout the day to ensure that the protocol continued to be met. The crushers were adjusted when necessary. The crusher quality control results were logged in a notebook, which was reviewed by AMEC. After each sample the crushers are cleaned out with compressed air.

Sample splitting is carried out in a Jones-type riffle splitter, with the riffle area dimensions of 28 cm x 9 cm, with gap widths of 15 mm. The riffle splitter and sample pans are cleaned with compressed air after each sample is processed. Rejects were placed in the original bags and the sub-samples are placed into 4” x 7” kraft tin-top sample bags provided by ALS Chemex. At the end of the shift some barren material (i.e. silica) is run through the crushers and the crushers are cleaned out with compressed air.

During the splitting the samples are split down to a sub-sample weight of approximately 200 to 300 g. The crushed sub-sample and remaining reject weights are recorded. Significant differences between the original dry weight and combined pulp and reject weight are compared to ensure that no sample mix-up occurred and as part of internal quality control. The Nevsun QAQC procedures require the insertion of 1 blank, 2 preparation laboratory pulp duplicates (two splits of a crushed sample), and 1 quartered core duplicate (described in Section 13.6)

The sample rejects are placed into large plastic drums, which hold about 20 to 25 samples. The drums are labelled as batches for all the samples done on that day. The prefix Bi-year-month-day (i.e. Bi-031027) is used to identify the batches. The samples are stored in six metre long shipping containers.

The lab is thoroughly cleaned at the end of each shift. The floors are swept and an industrial vacuum is used to clean equipment or the compressed air guns are used for difficult or inaccessible areas.

Data (sample numbers, weights) is entered in a spreadsheet to allow for quick retrieval and searches of data by lab staff or professionals on site.

AMEC consider the preparation laboratory and procedures in use to be acceptable and in accordance with standard industry practices.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-8
	October 2004

Sample Shipping

Groups of approximately 20 samples are packed in large plastic bags that are placed into the plastic shipping barrels. When samples are ready to be shipped the sample lists are combined with an ALS Chemex sample submission form and enclosed with the samples in the plastic drums. The lids of the plastic barrels are fastened with tamper proof zap straps. Samples are shipped once a week.

The sample information with required analytical procedures are emailed to ALS Chemex in Vancouver so that the sample shipments can be tracked and the Vancouver lab is made aware of the pending arrival of the samples.

Equipment Maintenance

Maintenance procedures were put in place in accordance to the equipment manuals. ALS Chemex laid out a daily, weekly and monthly maintenance schedule. A maintenance log was set up for entries of this maintenance.

13.4

Sample Preparation of RC Chips

Sample preparation of all RC chip samples was the same as the preparation described for the drill core samples.

13.5

Analyses

13.5.1

Genalysis Laboratory Services

All samples analyzed for gold at the Genalysis Laboratory by 50 g Fire Assay standard fusion method (Au by solvent extraction and flame AAS) with a 1 ppb detection limit²⁰ (Appendix D – Genalysis Analytical).

All samples analyzed for a 25 multi-element suite analysis used a 1 g aqua regia digestion, followed by Inductively Coupled Plasma (ICP-OES) analyses. The multi-element suite (with detection limits in parentheses) included: Ag (0.5 ppm), Al (20 ppm), As (2 ppm), Ba (2 ppm), Bi (2 ppm), Ca (0.01%), Cd (0.5 ppm), Co (1 ppm), Cr (2 ppm), Cu (1 ppm), Fe (0.01%), K (20 ppm), Mg (0.01%), Mn (1 ppm), Mo (2 ppm), Ni (1 ppm) P (20 ppm), Pb (2 ppm), S (10 ppm), Sb (10 ppm), Sc (1 ppm), Te (5 ppm), Ti (5 ppm), V (2 ppm) and Zn (1 ppm) (Appendix D – Genalysis Analytical). The aqua regia acid digestion is “total” for most base metals but is only “partial” for some of the major and minor elements.

²⁰ Mercier (2003) states that a detection limit has an uncertainty of +/- 100%. In other words, a detection limit of 1 ppb implies an uncertainty of 1ppb +/- 1ppb).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-9
	October 2004

Some limitations of the ICP-OES method were noted by Mercier (2003) as follows:

…spectroscopic techniques, such as ICP-OES, rely on being able to resolve the spectral signals unique to each element. When some of the signals are very strong (such as for the elements Fe and Al) they can interfere with the weaker signals from the other elements, and it may not be possible to achieve the optimum detection limits for these other elements. With the ICP-OES method problems are also encountered with elements, which are close to the detection limit.

A series of 25 stream sediment samples collected during the 2003 Phase I work by Mercier were also analyzed for platinum group elements (PGE’s). The method used 25 g Fire Assay Nickel Sulphide Collection followed by ICP-MS. The nickel sulphide button was pulverized and sample is digested with hydrochloric acid. The platinum group elements (with detection limits in parentheses) included: Ru (2 ppb), Rh (1 ppb), Pd (2 ppb), Os (2 ppb), Ir (2 ppb) and Pt (2 ppb) (Mercier, 2003).

13.5.2

ALS Chemex

All core and RC samples were sent to ALS Chemex for analyses. All samples were analyzed for gold by a 30 g fire assay fusion (Au AAS23) and determined analytically using an Atomic Absorption Spectroscopy (AAS) finish (Appendix D – ALS Chemex Analytical Procedures). Assays that were greater than the detection level (i.e. over limits) of the AAS finish (i.e. greater than 10,000 ppb) were re-assayed by a 30 g fire assay fusion (Au GRA21) and determined analytically using a gravimetric finish.

Multi-element analyses were completed with 41 elements Inductively Coupled Plasma - Atomic Emission Spectroscopy (ICP - AES) with Nitric-HCl Digestion (ME-ICP41A). This is the method used to determine the copper, zinc, lead and silver values (Appendix D – ALS Chemex Analytical Procedures). Copper, lead, zinc and silver samples that were greater than the detection level (i.e. over limits) of 50,000 ppm were re-assayed with Aqua Regia Digestion and Atomic Absorption Spectroscopy (AAS) (Appendix D – ALS Chemex Analytical Procedures). The 2002 drill core samples used the trace level ICP package (ME-ICP21) and were followed up with AAS for those samples that were over limits.

The few samples that were greater than the 30% detection level of AAS for base metals were assayed by wet assay titrimetric methods.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-10
	October 2004

Soil geochemical samples were tested using Inductively Coupled Plasma – Mass Spectroscopy (ICP-MS) to achieve ultra-trace detection levels on base metals and minor and major elements while gold determinations were completed with Inductively Coupled Plasma-Atomic Emission Spectroscopy (ICP - AES) on a fire assay fusion (Au-ICP21) for ultra-trace detection levels (Appendix D – ALS Chemex Analytical Procedures).

13.6

Nevsun Quality Assurance/Quality Control Program

Nevsun implemented Quality Assurance and Quality Control (QAQC) protocols for all exploration from the beginning of work on the Bisha Property, including work in 1998 and 1999. The QAQC samples used for geochemical sampling have little documentation and no presentation of the results or any corrective actions taken (if required). These samples were for the compilation of the exploration database and not part of the database for resource estimation.

All of the core and RC drilling programs included standards or certified reference materials (CRM). Drilling in 2003 and 2004 also included blanks, twin sample duplicates and coarse preparation duplicates. Each drill program report documented the protocols and results of the QAQC program. Nevsun did not do pulp duplicates and external check samples. AMEC recommends that approximately 5% of the sample pulps are re-submitted to a second laboratory as a check on the primary lab.

Table 13-1
Summary of Standards Used on the Drill Programs²¹

Standard Reference	Reference Material	Metal type	Mean	Unit	Standard Dev. (std.)	Mean + 2 std.	Mean - 2 std.
Standard A	GBM398-4C	Copper	3891	ppm	195	4281	3501
Standard A	GBM398-4C	Silver	48.7	ppm	5.1	58.9	38.5
Standard A	GBM398-4C	Zinc	5117	ppm	229	5575	4659
Standard A	GBM398-4C	Lead	11714	ppm	776	13266	10162
Standard B	G399-7	Gold	2660	ppb	120	2900	2420
Standard B	G399-6	Gold	2520	ppb	120	2760	2280
Standard C	GBM996-7C	Copper	2.35	%	0.145	2.64	2.06
Standard C	GBM996-7C	Silver	125.1	ppm	10.4	145.9	104.3
Standard C	GBM996-7C	Zinc	11.03	%	0.6	12.23	9.83
Standard C	GBM996-7C	Lead	3.89	%	0.278	4.44	3.33
Standard D	G999-4	Gold	4240	ppb	290	5400	4240
Standard E	GBM900-10	Copper	15.14	%	0.795	16.73	13.55
Standard E	GBM900-10	Silver	1549.6	ppm	75.6	1700.8	1398.4
Standard E	GBM900-10	Zinc	2.56	%	0.16	2.88	2.24
Standard E	GBM900-10	Lead	14.2	%	0.452	15.1	13.29
Standard F	G396-6	Gold	13.82	g/t	0.69	15.2	12.44
Standard F	G399-10	Gold	13.85	g/t	0.53	14.91	12.79
Standard F	G900-2	Gold	1.48	g/t	0.06	1.6	1.36

²¹ 2002 standards not included in the list

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-11
	October 2004

Nevsun purchased the standards from Geostat - Sample and Assay Monitoring Service, located in Australia. The nine standards used over the duration of the drill programs, since 2003 are shown in Table 13-1. The reference material includes a range of low-grade, mid grade and high-grade precious and base metal standards with certified values and statistically acceptable limits. The maximum and minimum limits used by Nevsun for monitoring are plus two standard deviations (+2st.dev.) and minus two standard deviations (-2st.dev.) from the mean value of the control sample.

The Nevsun QAQC sampling protocols and standard insertion instructions are provided in Appendix D. The QAQC sample insertion protocol employed by Nevsun for all core and RC drill sampling subsequent to the 2002 program includes the following samples:

• 6 certified standard control samples per 100 samples; 3 gold (B, D, and F) and 3 base metal (A, C, and E).
  

• 1 coarse blank sample of barren material per 100 samples; as well as, barren material randomly inserted in mineralized zones.
  

• 1 quartered core “twin” duplicate sample per 100 samples.
  

• 2 coarse preparation duplicates per 100 samples.
  

The QAQC program for the 2002 drilling included 11 insertions of a standard (denoted CRM on Table 13-2). Four of the 11 insertions were not within the accepted limits. These standards were not used for subsequent sampling programs after significant mineralization encountered at Bisha demanded a more substantial and thorough QAQC program.

During the 2003 Phase I and II and 2004 drilling programs, a total of 1,310 insertions of standards were made into the sample sequence of 20,545 core and RC samples (Table 13-2). In addition to the standards were 352 blanks, 225 twin duplicates and 372 coarse preparation duplicates. In total the QAQC samples comprise 11% of the total sample analyses.

A total of 83 of the standard sample insertions were not within the allowable limits. Of these 83 samples, 51 samples were resolved by re-assaying the entire sample batch or re-assaying those samples within the batch that were at the same relative grade range and/or same analytical method. Of the 32 samples that remained unresolved, 15 samples were reported by the assay lab to have “not sufficient sample” (NSS) for a re-assay to be completed. Standards E and F had 8 samples that were not re-assayed because the metal grades for the surrounding samples were low grade and the standards were high and therefore used a different analytical method. The remaining 9 samples that were not resolved were due to being either sample mix-ups or contamination of the sample.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-12
	October 2004

Table 13-2
Summary of Standard for Core and RC Sampling

	2002			2003 Phase I			2003 Phase II			2004
Standard	Samples	Re-Assay	Resolved	Samples	Re-Assay	Resolved	Samples	Re-Assay	Resolved	Samples	Re-Assay	Resolved
A				30	2	1	46	7	4	148	7	7
B				34	0	0	43	5	4	143	5	1
C				32	1	1	41	2	2	139	0	0
D				32	0	0	42	4	4	145	2	0
E				31	8	2	41	11	9	144	11	9
F				30	3	1	40	5	5	138	6	1
CMR	11	4	0
Totals	11	4	0	189	14	5	253	34	28	857	31	18

The coarse blank material was sourced from near the Bisha Property. This material usually consists of limestone and/or dolomite. Although the samples were considered to consist of barren rock without any appreciable precious metal or base metal content, occasional low levels of mineralization could conceivably occur within the material and therefore would negate the usefulness of this material as a blank. Table 13-3 shows a large amount of the metal values that were returned at greater than the 3 times the detection limit of the ICP analytical method. After cursory review of the logs, sample batches, and data it is apparent that the blank material is not barren and thus the true values of the blank for each metal is not known.

During the review of the data for Cu and Au it is also clear that cross-over contamination has occurred for some intervals within mineralization. There are 55 copper values (of the 249 values in Table 13-3) and 23 gold values (of the 47 values in Table 13-3) that show possible contamination from previous higher-grade samples. All the copper and 20 of the gold values are from within the massive sulphide mineralization.

Table 13-3
Summary of Blanks Greater than the 3x Detection Level

	Number Metal Values > 3x Detection Level
Year	Au	Ag	Cu	Pb	Zn
2003 - I	8	3	23	2	13
2003 - II	7	1	30	7	17
2004	32	19	196	54	131
Total	47	23	249	63	161

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-13
	October 2004

A series of Quarter core twin samples of the diamond drill core were prepared and submitted. Two samples for the same sample interval were prepared by cutting the core in half and then cutting the core into two “quarter” core samples. Differences between the sample values are considered to reflect the inherent nugget effect of the mineralization. The evaluation of the sample pairs used a failure boundary corresponding to a 30% relative error. The analysis of 225 twin samples (1.1% of the total number of samples) yielded 63 sample pairs outside of the 30% relative error for Au, Ag, Cu, Pb or Zn (see Figure 13-3), representing 28% failures. However after applying a lower cut-off to each metal of 0.1 g/t to Au, 10 g/t to Ag, 0.05% to Pb and 0.25% to both Cu and Zn there was a reduction of failures to 20 sample pairs. The cut-off is applied because it is unrealistic to consider the ±30% relative error for samples with values close to the detection limits.

After exclusion of the sample pairs below the cut-off, the proportion of sample pairs above the 30% relative error is 8.9%. Ag has 5 sample pair failures (or 2.6%) and Au had 5 sample pair failures (or 2.2%). AMEC considers the sampling variance to be within an acceptable range.

Coarse preparation duplicates provide data on the precision or homogeneity of the sample after being crushed and split. Nevsun did not collect coarse duplicates until the 2003 Phase II drill program and commencement of the on-site preparation facility. The evaluation of the sample pairs used a failure boundary corresponding to a 20% relative error. The analysis of 372 coarse duplicates (1.8% of the total number of samples) yielded 76 sample pairs outside of the 20% relative error for Au, Ag, Cu, Pb and Zn (see Figure 13-4), representing 20.4% failures. However after applying a lower cut-off to each metal of 0.05 g/t to Au, 10 g/t to Ag, 0.05% to Pb and 0.25% to both Cu and Zn there was a reduction of failures to 33 sample pairs.

Au and Ag have the largest set of sample pair failures with Au having 13 and Ag having 11 or 3.5% and 2.9% respectively. Excluding the sample pairs below cut-off, the impact of the failing pairs for Au, Ag, Cu, Pb and Zn is reduced to a reasonable 8.9%. AMEC concluded that the sub-sampling variance for the studied elements (Au, Ag, Cu, Pb and Zn) was within acceptable ranges.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-14
	October 2004

Figure 13-3
Graphs for Twin Samples for Gold, Silver, Copper, Lead and Zinc

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-15
	October 2004

Figure 13-4
Graphs for Coarse Preparation Duplicates for Gold, Silver, Copper, Lead and Zinc

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-16
	October 2004

13.7

Security

The chain-of-custody for core samples collected and being shipped from site is as follows:

• Core is transported to the Bisha camp by the drill contractors and placed in the core logging area.
  
• Logging and sample preparation area and Bisha camp is a fenced and guarded compound.
  
• Core samples are crushed and sub-sampled (see Section 13.3.3).
  
• Prepared samples are placed in sealed barrels.
  
• Each barrel has a list of samples written on the outside of the container.
  
• A sample submission form accompanies each barrel.
  
• Barrels are transported to Asmara in company-owned vehicles arranged by Nevsun.
  

The sample barrels are submitted to the Ministry of Mines for inspection and submission to customs, a customs seal is placed on the barrels and they are shipped via air transport directly to ALS Chemex in Vancouver, Canada.

AMEC considers the security and chain-of-custody procedures to be reasonable and acceptable.

AMEC accompanied the indpendent samples that were collected and prepared as part of this study from the preparation lab to the Ministry of Mines. The renumbered and randomized sample sequence would prevent any systematic tampering with the samples.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 13-17
	October 2004

14.0

DATA VERIFICATION

14.1

Data Verification by Nevsun

Data verification activities by Nevsun included:

• Visual checks of entered data.
  

• Periodically checking the database for extreme values or codes that were not within the accepted set of codes.
  

• Importation of the drill hole database to Gemcom for validation of hole lengths and identification of overlapping geological and assay intervals.
  

All problems identified by Nevsun were resolved as they were encountered resulting in a final database for submission to AMEC for auditing.

Assays were imported directly from digital laboratory certificates thereby minimizing the opportunity for entry errors of assay data.

No double entry, pick lists or filtering of data was in use during entry.

14.2

Data Verification by AMEC

During the Bisha site visit AMEC reviewed the available drilling and other exploration and project data. A database with a total of 288 diamond drill holes with a cumulative meterage of 45,216 m was available for review but the collar survey and assay portions of the database were incomplete. A total of 40 RC drill holes were recorded in the database, however collar surveys, assays and other information was incomplete.

The data review conducted on site included:

• Core hole database review: collars, surveys, lithology, minor-litho, alteration, mineralization, structure, and available assays.
  

• RC hole database review: collars, surveys, alteration, mineralization, structure, and lithology. No assays were available.
  

• Comparison of drill hole collar surveys to locations on topographic maps.
  

• Resurvey of drill hole collars (6 holes).
  

• Downhole survey review and readings of Sperry Sun™ disks.
  

All problems or errors encountered during the site work were documented and provided to Nevsun for checking or correction.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-1
	October 2004

Other review and verification activities completed by AMEC on site included:

• Core logging review.
  

• Core recovery and RQD review.
  

• Core facility, cutting facility, sample preparation facility, and storage area reviews.
  

• Collection of quartered core samples (40).
  

• Inspection of sample preparation equipment.
  

• Quality control checks on sample preparation (20 sieve checks).
  

• Selected 40 rejects from storage for sieve checks (20) and sub-samples (40) for check assays.
  

• QAQC checks: Standards, Blanks, Preparation duplicates (20 from the first split, 20 from the last split).
  

• Samples collected by AMEC were renumbered, randomized and submitted “blind”.
  

• Transported 172 samples from site in AMEC custody.
  

• Submitted the 172 samples to the Ministry of Mines for customs inspection and shipping to ALS Chemex Laboratory in Vancouver, Canada.
  

During the site work AMEC visited the Main Zone gossan, Northwest Zone, Guardian Hill, and Conical Hill. Field observations were compared to available maps and interpretations.

After the drill hole program was finished and the data was compiled and verified by Nevsun, a final database review was completed by AMEC.

Additional verification activities included checking high values and relationships between grades and sample lengths. High values for Au, Ag, Cu, Pb, and Zn for each rock type were investigated and checked to confirm that the logged mineralization did concur with the assay results. Several intervals in question were not within the geologic model and therefore these intervals were not investigated further. Nevsun should check these intervals prior to the next resource estimate.

Potential relationships between grade and sample lengths were investigated and no direct correlation was observed.

Nevsun was advised of all problems or inconsistencies that were noted during the AMEC’s review and Nevsun rectified these items. AMEC considers the final database to be robust and verified.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-2
	October 2004

14.3

AMEC Quality Control Checks

To check on the adherence of the sample crushing procedures, AMEC conducted a series of sieve checks on current samples (Table 14-1) and also on reject material of samples that were pulled from storage (Tables 14-1 and 14-2).

Table 14-1
Sieve Checks for Samples from 2004 Program

Sample #	Passes through Splitter	Rock Code	Total Weight (g)	< 2 mm (g)	% passing 2 mm
335523	4	SOAP	0.220	0.164	75%
335543	3	SOAP	0.368	0.282	77%
335562	3	SOAP	0.224	0.172	77%
335561	4	SOAP	0.230	0.198	86%
335317	4	SOAP	0.184	0.116	63%
335573	4	MAFT	0.246	0.154	63%
335565	4	MAFT	0.348	0.260	75%
335567	4	MAFT	0.342	0.256	75%
335566	3	MAFT	0.318	0.238	75%
335574	3	MAFT	0.244	0.164	67%
335509	4	OXID	0.260	0.198	76%
335506	4	OXID	0.282	0.234	83%
335507	4	OXID	0.220	0.172	78%
335505	4	OXID	0.250	0.192	77%
335502	4	OXID	0.230	0.178	77%
335558	4	SUIP	0.450	0.420	93%
335264	3	SUIP	0.268	0.188	70%
335559	3	SUIP	0.512	0.496	97%
335555	4	SUIP	0.424	0.318	75%
335550	4	SUIP	0.470	0.406	86%
Average					77%

The current sample preparation is within the accepted protocol of 70% passing 2 mm. AMEC noted that the sample preparation personnel regular check that the crushed material is meeting the protocol.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-3
	October 2004

Table 14-2
Sieve Checks for Samples from Pre-2004 Programs

Sample #	Passes through Splitter	Total Weight (g)	< 2 mm (g)	% passing 2 mm
280959	3	0.634	0.372	59%
280962	3	0.588	0.352	59%
280965	3	0.574	0.346	59%
286177	4	0.650	0.446	65%
286251	3	0.322	0.248	32%
286259	4	0.754	0.508	75%
286407	4	0.366	0.312	36%
286413	5	0.396	0.294	39%
291841	3	0.386	0.298	38%
334176	3	0.310	0.234	31%
Average				49%

Sieve checks on sample material that was pulled from storage was variable and many samples did not meet the current protocol (Table 14-2). A subsequent check of the original assay versus an assay of the reject material (Section 14.4.2) showed relatively good agreement (most samples within ±20%) and therefore AMEC does not consider this to be of concern.

As a check on the quality of the data entry, AMEC completed a small double data entry check. Of 480 records entered, there were over 20% discrepancies but this was due to a change in lithologic and mineralization coding and additional detail of mineralized intervals that was added in the holes selected. The hard copy and digital information should match therefore a revision of the hard copy logs is advisable.

AMEC recommends the use of either a double entry system or a data entry system with some form of validation of codes. Direct entry into MS Access or some other relational database with filters, limits, and data integrity checks could be implemented.

14.4

AMEC Independent Sampling

AMEC collected a series of samples during the site visit, as noted in Section 14.2. A total of 172 samples were submitted to ALS Chemex for analyses (Table 14-3, Appendix B). The results for the standards were reasonable (Section 14.4.3) and AMEC did not submit a subset of samples to a second laboratory.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-4
	October 2004

Table 14-3
Independent and QAQC Sampling

Type	Number
quartered core sample	42
sub-samples of reject material	40
standards (4 of each of the 6 standards)	24
blank (coarse blank material)	6
“first” splits from new samples	30
“last” splits from new samples	30
Total	172

AMEC conducted or was present during the collection and preparation of the samples. The samples were placed in a randomized sample sequence and renumbered which would prevent any systematic tampering with the samples. AMEC accompanied the samples from the preparation lab to the Ministry of Mines office in Asmara.

No outcrop samples were collected due to the extreme oxidized and variable nature of mineralization. AMEC did not consider that resampling surface samples would provide any reasonable comparisons. AMEC observed drilling in progress and is confident of the presence of base metal mineralization and concludes that the samples of quarter core and rejects provide confirmation of the grades and reproducibility of assay values.

14.4.1

Quartered Core

Forty-two samples of quartered core were collected from 9 holes (B-9, B-14, B-17, B-20, B-21, B-24, B-43, B-126, B-140; Photos A-31 to 35 in Appendix A). Two of the samples were clearly swapped (sample numbers 338070 and 338056) and graphs for Au, Ag, Cu and Zn are provided in Figure 14-1. Of the remaining 40 sample pairs there were four pairs that plotted outside of the +/-30% limits for Au, three for Ag, none for Cu, and one for Zn.

Comparisons of half-core to quartered core are difficult due to the change in size of sample. However, AMEC considers these samples to show a reasonable reproducibility.

14.4.2

Sub-sampling of Reject Material

Reject material for 40 samples that were collected and prepared during previous drilling programs were removed from storage and submitted for assay. The samples were from nine drill holes that were part of the 2003 and 2004 drilling programs. Graphs are provided in Figure 14-2.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-5
	October 2004

Figure 14-1
Original versus Quarter Core Sample Pairs

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-6
	October 2004

Figure 14-2
Original versus Reject Sample Pairs

Of the 40 sample pairs, one pair plotted outside of the ±20% limits for Au, three for Ag, two for Cu, and four for Zn (Figure 14-2). AMEC considers these results to show a reasonable reproducibility and provides assurance that the sample homogenization prior to splitting is reasonable.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-7
	October 2004

14.4.3

Splits Versus Original Sample

As a further test of the sample homogenization during the sample preparation, AMEC collected the first and last splits that are normally rejected during the sub-sampling using the Jones Splitter. Thirty samples were processed and of these pairs, three plotted outside of the ±20% limits for Au, one for Ag, three for Cu and one for Zn (see Figure 14-3). AMEC expects at least 90% of the samples to fall within the limits (±20%) therefore this set of sample pair is accepted. AMEC considers, however that the need for ensuring that the crushing protocols are being met is underscored.

Figure 14-3
Original vesus Last Split Sample Pairs

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-8
	October 2004

The first splits were also compared to the original sample and were found to have similar results for the sample pairs (results included in Appendix B; graphs are not provided in this report).

14.4.4

Standards

Nevsun used six commercial certified standards during the submission of samples (see Section 13.9). Standards A, C and E are base metal and silver standards; standards B, D and F are gold standards only. AMEC renumbered and repackaged four portions of each of the six standards and submitted them into the sample sequence as blind standards.

Table 14-4
Standard Samples Submitted with AMEC Samples

Standard Samples Submitted with AMEC Samples						Certified Values for Standards						Difference
Sample	Au g/t	Ag g/t	% Cu	% Pb	% Zn	Sample	Au g/t	Ag g/t	% Cu	% Pb	% Zn	Au g/t	Ag g/t	% Cu	% Pb	% Zn
338165	0.159	51	0.43	1.10	0.48	A		48.7	0.39	1.17	0.51		5%	9%	-7%	-6%
338139	0.149	49	0.40	1.14	0.48	A		48.7	0.39	1.17	0.51		1%	2%	-3%	-7%
338135	0.139	49	0.40	1.12	0.49	A		48.7	0.39	1.17	0.51		1%	2%	-4%	-5%
338044	0.17	51	0.40	1.16	0.49	A		48.7	0.39	1.17	0.51		5%	2%	-1%	-4%
338117	2.54	4	0.06	0.02	0.07	B	2.52					1%
338013	2.35	4	0.08	0.02	0.07	B	2.52					-7%
338145	2.54	4	0.06	0.02	0.08	B	2.52					1%
338148	2.53	5	0.07	0.02	0.08	B	2.52					0%
338172	0.385	123	2.28	3.71	10.70	C		125.1	2.35	3.89	11.03		-2%	-3%	-5%	-3%
338107	0.424	123	2.13	3.59	10.70	C		125.1	2.35	3.89	11.03		-2%	-9%	-8%	-3%
338010	0.427	127	2.34	3.87	10.85	C		125.1	2.35	3.89	11.03		2%	0%	-1%	-2%
338143	0.441	123	2.40	3.97	10.95	C		125.1	2.35	3.89	11.03		-2%	2%	2%	-1%
338168	4.86	2	0.27	0.03	0.04	D	4.24					15%
338060	4.61	<1	0.26	0.03	0.04	D	4.24					9%
338098	4.66	3	0.27	0.04	0.03	D	4.24					10%
338004	4.52	2	0.26	0.03	0.04	D	4.24					7%
338102	16.7	1405	16.40	14.55	2.58	E		1549.6	15.14	14.20	2.56		-9%	8%	2%	1%
338041	16.95	1435	1.59	14.55	2.54	E		1549.6	15.14	14.20	2.56		-7%	-89%	2%	-1%
338034	16.9	1490	1.58	14.65	2.56	E		1549.6	15.14	14.20	2.56		-4%	-90%	3%	0%
338096	16.15	1455	15.90	14.50	2.53	E		1549.6	15.14	14.20	2.56		-6%	5%	2%	-1%
338063	13.3	5	0.02	0.01	0.01	F	13.85					-4%
338092	13.5	4	0.01	0.01	0.01	F	13.85					-3%
338147	13.75	6	0.02	0.01	0.04	F	13.85					-1%
338059	9.59	6	0.02	0.01	0.06	F	13.85					-31%

The data set is small but the results are good with few exceptions (Table 14-4). Samples 338041 and 338034 are significantly lower than the certified standard value for Cu. Both of these samples were assayed by ALS Chemex as overlimit samples and AMEC considers that the laboratory erroneously reported these two values with the decimal place in the wrong position. The last sample (338059) returned an incorrect assay for Au; samples for standard D are variable and consistently are higher value but are all within the limits (±2 Standard Deviations; see Table 13-1).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-9
	October 2004

14.4.5

Blanks

AMEC included six samples of the same coarse blank material as used by Nevsun. Unfortunately this material is a local rock, which cannot be certified to be sterile. Assay results for these samples were all very low but many of the individual values are above the accepted limit of three times the detection limit.

Table 14-5
Blank Samples Submitted with AMEC Samples

Type	Sample	Au g/t	Ag g/t	% Cu	% Pb	% Zn	Certificate	Date
Blank	338043	0.010	1	0.0097	0.009	0.005	VA04034670	07-Jun-04
Blank	338126	0.076	1	0.0023	0.003	0.002	VA04034670	07-Jun-04
Blank	338065	0.022	1	0.0123	0.008	0.040	VA04034670	07-Jun-04
Blank	338048	0.022	1	0.0067	0.004	0.023	VA04034670	07-Jun-04
Blank	338095	0.015	1	0.0053	0.001	0.005	VA04034670	07-Jun-04
Blank	338003	0.014	1	0.0158	0.001	0.157	VA04034670	07-Jun-04
Blank	-	0.005	1	0.0005	0.001	0.001	Detection limits

AMEC recommends that Nevsun purchase a commercial blank for use. If the use of a coarse blank material is continued then it should be of a clean, barren material with no obvious oxidation surfaces or patches of limonite, hematite, etc.

14.4.6

Bulk Density Checks

A suite of 40 samples was sent to PRA (Process Research Associates) for dry bulk density measurements using both wax and without wax immersion methods. The moisture content was also measured for these samples. The procedures used and the results for the measurements are provided in Appendix B.

The main purpose of the measurements was to examine the difference between wax and non-wax measurements (Table 14-6). The differences average –1% but range from +3% to -6% relative to the wax immersion method. For the majority of the samples the differences are minor. The use of bulk densities is reviewed in Section 11.2.7 of this report.

Grouping of this small set of bulk density samples by mineralized domains shows average values that are actually higher than the accepted (see Table 14-7 and Section 11.2.7). Host rocks have an average bulk density of 2.83 based on 18 samples.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-10
	October 2004

Table 14-6
Bulk Density Measurements using Wax and Non-Wax Immersion Methods

Sample ID	No WAX Bulk Density g/cm3	WAX Bulk Density g/cm3	Difference %	Moisture %
B-232 85.5m SMSX MAFT	3.83	3.72	-3%	0.03
B-232 56.5m STSX (MAFT)	2.82	2.67	-6%	0.12
B-233 42.15m (MAFT)	2.61	2.69	3%	0.15
B-233 MSUL 62m (PRIM)	3.93	3.85	-2%	0.04
B-233 FELD 23.2m	2.47	2.47	0%	0.13
B-233 STSX (XTMF) 212m	2.77	2.82	2%	0.02
B-236 SOAP 70.1m	2.48	2.53	2%	0.35
B-242 104.7m MAFD	2.72	2.67	-2%	0.15
B-242 25.3m MAFT	3.01	2.99	-1%	0.20
B-242 68m SOAP/MAFT	2.34	2.20	-6%	4.82
B-242 84m MSUL/SUPL	4.48	4.48	0%	0.05
B-242 128m MSUL/PRIM	4.21	4.16	-1%	0.00
B-243 FELD 42.2m with STSX	2.70	2.70	0%	0.04
B-243 MAFT 92m	2.82	2.70	-5%	0.07
B-243 FELD 132m	2.72	2.78	2%	0.07
B-243 CTBX 170.85m	2.56	2.58	1%	0.23
B-243 MSUL (PRM) 187.5m	4.13	4.19	1%	0.05
B-243 SYSX 209.3m	3.49	3.45	-1%	0.03
B-243 STSX (MAGT) 219.3m	2.67	2.71	2%	0.03
B-246 70m MAFT	2.72	2.71	0%	0.05
B-254 45.5m MAFT	2.68	2.59	-4%	0.36
B-254 93m MSUL (SUPG)	4.49	4.61	3%	0.01
B-254 138m MSUL (PRIM)	4.77	4.82	1%	0.06
B-254 146m MSUL (PRIM)	4.85	4.88	1%	0.00
B-256 76m MSUL (SUPG)	4.67	4.64	-1%	0.00
B-256 116m MSUL (PRIM)	4.87	4.85	0%	0.00
B-256 145m STSX (MAFT)	2.78	2.65	-5%	0.04
B-258 66.9m FPDK (OXID)	1.52	1.50	-2%	7.07
B-258 84m SOAP (ACID)	2.21	2.11	-5%	1.23
B-258 104.5m STSX (MAFT)	2.86	2.84	-1%	0.01
B-258 22.2m FERC (OXID)	3.64	3.65	0%	0.23
B-258 38.0m FERU (OXID)	3.99	3.97	-1%	0.15
B-259 68.0m MSUL (SUPG)	3.96	3.89	-2%	2.98
B-259 106.0m SMSX (MAFT)	2.73	2.80	2%	0.12
B-260 55.7m MSUL/SUPG	4.37	4.39	1%	0.01
B-260 120m MSUL/PRIM	4.43	4.45	0%	0.02
B-262 79.0m MSUL/SUPG	4.46	4.46	0%	0.23
B-272 83m MSUL/SUPG	4.53	4.44	-2%	0.00
B-272 103m MSUL/PRIM	4.66	4.55	-2%	0.00
B-272 113m STSX/MAFT	2.89	2.92	1%	0.04
Average	3.42	3.40	-1%	0.48
Minimum	1.52	1.50	-6%	0.00
Maximum	4.87	4.88	3%	7.07

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-11
	October 2004

Table 14-7
Bulk Density Measurements using Wax Immersion Methods

Domain	Code	Count	Average Bulk Density g/cm3	Minimum Bulk Density g/cm3	Maximum Bulk Density g/cm3
Breccia	BX	1	2.58	-	-
Oxide	OXID	2	3.81	3.65	3.97
Acid	ACID	3	2.28	2.11	2.53
Supergene	SUPG	6	4.41	3.89	4.64
PrimaryZn	PRIM	9	4.47	3.85	4.88
Felsic Dyke	FPDK	1	1.5	-	-
Host Rock (Volcanics)	MAFT	18	2.83	2.47	3.72

The moisture values for these same samples were low, averaging 0.48% and ranging from 0.0 to 7.07% (Table 14-6). The highest values were in SOAP and OXID units.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 14-12
	October 2004

15.0

ADJACENT PROPERTIES

Adjacent properties include the Augaro, Okreb and AK Properties located to the south, east and north of the Bisha property respectively. These properties are all held by Nevsun and are the subject of upcoming exploration programs. Sanu Resources and other companies hold other nearby Exploration Licenses but no details were available to AMEC regarding prospects or mineralization on those properties.

The Augaro gold mine was exploited during the Italian colonial times (see Section 7.3). Few details are available regarding production or geology. The mineralization is a shear zone hosted gold deposit (pers. comm. Nielsen, 2004) with a surficial gossan and gold mineralization.

The Okreb Properties are now part of the Bisha Exploration License. Some drilling, geophysics and geochemical sampling have been completed. No specific deposits have been identified to date.

Few details are available for the AK Property.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 15-1
	October 2004

16.0

MINERAL PROCESSING AND METALLURGICAL TESTING

Lynton Gormely, P.Eng., Principal Process Engineer (AMEC Vancouver office) completed a review of available data and information on metallurgical testing completed by Process Research Associates Limited (PRA) on samples collected and submitted by Nevsun. PRA is a metallurgical and process testing laboratory located in Vancouver. The testwork included cyanidation of oxide material and flotation tests of supergene copper enriched material from crushed rejects (-10 mesh) from drill core samples (Appendix E).

16.1

Cyanidation

The samples submitted for cyanidation were core of the high grade oxide intervals in holes B-20 and B-38. The samples had gold grades of 23.3 and 10.8 g/t Au respectively. High recoveries are necessary to produce acceptable (low value) residues from high grade feeds.

Diagnostic leach tests suggest that 90 to 95% of the gold is directly cyanide soluble. In 72 hours cyanidation (PRA test C3), the sample from hole B-20 yielded 96.1% gold recovery, or a residue of 0.91 g/t Au, which might be judged as acceptable, although there is still an economically significant amount of gold remaining in the residue. Silver in the residue is 10.7 g/t.

In 72 hours cyanidation (PRA test C4), the sample from hole B-38 yielded 82.1% recovery, or a residue of 1.97 g/t Au. This residue contains an even higher gold content.

Cyanide consumptions for the two tests were low: 0.17 and 0.33 kg/t, respectively. Lime consumptions were low to moderate: 1.78 and 3.73 kg/t, respectively.

More work is needed to be able to optimize the cyanidation of the high grade oxide feed in order to reduce the grades of the residues to insignificant precious metal contents.

Possible options to improve recoveries include: ultra fine grinding, intensive cyanidation or a combination of these methods. Gold is reportedly micron size and therefore gravity processing of the material may not be feasible.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 16-1
	October 2004

16.2

Flotation

The core samples submitted for flotation tests were from the supergene copper intervals in holes B-41 and B-4. The testwork was completed to examine the potential for upgrading the copper content by flotation. Detailed test results and head grades were not provided; however, the samples for hole B-41 that were used for testing have an average grade of 8% Cu and the samples for hole B-4 that were used for testing have an average grade of 3.8% Cu.

The concentrate grades that were achieved were not upgraded significantly from the (possible) head grades of the samples, as estimated above. Flotation may not have been successful in upgrading the samples due to the high pyrite content (~95%) in the products. The mineralogical report for the samples identified high amounts of pyrite.

Four additional samples collected from the massive sulphide component of the resource were tested for upgrading by flotation. Copper recoveries of the samples ranged from 79 to 94% in 8.5 to 19% of the feed. Gold and silver recoveries for these samples through flotation were poor at 38 to 51% Au recovery and 53 to 77% Ag recovery.

16.3

Conclusion

The testwork completed to date is to a scoping level of investigation, which is sufficient to provide guidance for more definitive research. Insufficient work has been done to support determination of engineering design criteria and economic indicators. More testing is required, but based on the material reviewed by AMEC, it is reasonable to expect that suitable metallurgical procedures for economic recovery of the metals will be identified through further study.

The deposit mineralogy is polymetallic and covers a full spectrum of the degree of oxidation of the mineralizatin and will require a significant testwork program to identify an optimum process flowsheet and to confirm metal products, grades and recoveries. A process design will likely include several process recovery circuits – i.e. sulphide flotation with several circuits (i.e. Cu and Zn banks) followed by cyanidation of flotation tails.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 16-2
	October 2004

17.0

MINERAL RESOURCE AND MINERAL RESERVE ESTIMATES

17.1

Current Work

The 2004 resource model for the Bisha project contains blocks classified as Indicated and Inferred mineral resources for the Bisha Main Zone. Grades have been estimated beyond the limits of these classifications. The unclassified blocks have no resource classification and may be used for such activities as developing drill targets for development of the feasibility level resource model.

The resource model deals with the main zones of the Bisha Deposit. This model does not include grade estimation for ‘stringer’ zones (on the footwall side of the massive sulphide deposit) and at least two other smaller mineralized zones that appear to carry potentially economic grades. The drill hole data for these zones is in the existing database. These zones should be modelled in either an update to this model or in the feasibility level resource model.

This mineral resource estimate was completed using Vulcan® mine modelling software by Ken Brisebois, P.Eng., Senior Geologist with AMEC (Phoenix office).

17.2

Summary, Conclusions and Recommendations

17.2.1

Resource Database and Geological Models

The current database is adequate for preparation of a long range model for pre-feasibility modelling of the deposit. The deposit is generally drilled on nominal 25 m spacing between sections and with variable (but mostly 25 m) spacing on sections. The geological modelling work was completed in Gemcom modelling software, prior to importation into the Vulcan software.

The lithologic domains interpreted and modelled in 3D at Bisha serve to adequately domain the deposit for this level of investigation. Given that there are several metals to estimate, one could consider optimizing the model by interpreting different domains for different metals. This approach should be considered for the feasibility level resource model.

17.2.2

Summary of Ore Controls

The modelled domains provide a reasonable global framework to limit the estimation and extrapolation of economic gold and base metal mineralization within the deposit. Exploratory data analysis (EDA) of assays and composites within and outside the domain boundaries reveal several pertinent issues:

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-1
	October 2004

• There are different trends in the various metals with respect to the modelled lithologic domains. To both a lesser and greater degree, each metal exhibits unique trends within the deposit.
  

• The distributions are generally well-behaved with relatively low coefficients of variation with some notable exceptions such as Au and Ag in the Acid domain.
  

• Grade profiles for each metal across the various domain boundaries illustrate a variety of behaviours ranging from soft (little change in average grade) to hard boundaries (abrupt change in average grade very close to the contact).
  

• Analyses of correlations between the various metals show complex relationships.
  

To address these issues, the following concepts were developed to assist in domaining the deposit for grade estimation:

• The outer boundaries of the lithologic domains were used to form hard boundaries to separate the potentially economic mineralized material from the largely un-mineralized or sub-economic material outside of the interpreted domains.
  

• There was no estimation of grade outside of the lithologic domains.
  

• Within the lithologic domains, and groupings of these domains identified during the EDA phase of the work, ordinary kriging was determined to be a suitable estimator for this level of investigation.
  

• Correlations between various metals, while complex, should be further studied in the next resource model. The current model makes a reasonable effort to reproduce the correlations between the various metals but could likely be improved. An improvement in this area would be unlikely to affect the global resource inventory but may improve local estimation in the model.
  

17.2.3

Grade Capping

Grade capping is used in this resource model to limit the spatial extrapolation of the occasional, but generally isolated, anomalously high grades. For this study, AMEC has utilized the AMEC simulation-based approach²². Results were compared to subjective cap grade levels from probability plots.

Using these inputs and the declustered assay distributions, a percentage metal-at-risk was calculated for each metal to be estimated. The results are presented in Table 17-4. The metal-at-risk values are typical for this style of deposit and at this stage of project development. The levels will likely decline somewhat as drilling continues on the property.

²² Developed by MRDI, a predecessor of AMEC Mining & Metals.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-2
	October 2004

Ag is problematic but the very high grades are ma'inly restricted to the Acid domain, which is volumetrically contained and relatively small. The metal-at-risk amounts were removed during the estimation process.

17.2.4

Variography

Variograms were calculated for the five metals within each of the estimation domains. The variogram modelling was completed within the SAGE 2000 variogram software using the 3 m composite database.

At Bisha, spatial variability is complicated by local trends of high and low-grade mineralization. Some of the trends are naturally occurring. Some of these trends emerge from the necessity of combining populations where there is not enough data to support two sub-populations. Some of the poor correlation is caused, for example, by the isolation of the Acid domain. This is more a practical necessity rather than a geostatistical decision due to its anomalous nature in a comparatively small volume. All of these factors have led to variograms that, in a few cases, exhibit relatively poor structure. In most cases, variograms are moderate to good leading to relatively confident estimation results.

17.2.5

Block Model Validation

To validate the kriged estimates, a number of checks were carried out. These included:

• Visual inspection of kriged results on plans and sections.
  

• Comparison of kriged and nearest neighbour statistics.
  

• A change of support analysis.
  

The results of checks validated the model results adequately.

Change of support analysis was completed for the principle mineralized zones. This analysis highlighted some areas for improvement in the subsequent resource models. These improvements, while important for the upcoming mine planning, are not expected to result in a significant change in the global metal content for the resource and reserve.

Internal dilution of the block grades is deemed to have been addressed by the use of the lithologic domains and tagging of all of the drill hole sampling within the domains. All samples within the boundaries of the domains, whether below economic cut-offs or not, have been used to estimate block grades.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-3
	October 2004

Contact dilution has not been addressed in this model however. Blocks that occur along the edge of the domains have been estimated with composites occurring only within the respective domains. There is, however, a percentage of mineralization stored in each block of the model such that the in-situ resource can be tabulated. It is recommended that contact dilution be further addressed during the feasibility level studies. Two or three areas should be chosen where the hanging wall and footwall could be drilled on a detailed spacing such as 12.5 m on section and 5 m in the vertical. These data may then be used to analyze the deposit on a scale whereby applicable dilution factors could be developed.

17.2.6

Resource Classification

The mineral resources have been classified according to guidelines and logic summarized within the CIM Definitions referred to in National Instrument 43-101. Resources were classified as Indicated or Inferred by considering geology, sampling and grade estimation aspects of the model. For geology, consideration was given to the confidence in the interpretation of the lithologic domain boundaries and geometry. For sampling, consideration was given to the number and spacing of composites, the orientation of drilling and the reliability of sampling. For estimation, consideration was given to how well the kriged grades reflect the composites and how high-grade composites were projected into plan and section. For the estimation results, consideration was given to the confidence with which grades were estimated.

Based on the results of confidence interval calculation for the block grade estimations, the following points relate to the resource classification criteria:

• Confidence interval calculations would suggest that 25 to 30 m spaced sampling is sufficient to define an Indicated Resource.
  

• In some cases, such as Cu in the Supergene and Primary domains, the ability to estimate the Cu grade with reasonable confidence is more than 25 m spacing.
  

• In the case of Au in the Acid domain, confidence in estimation is quite low.
  

Synthesizing the confidence analyses as well as the other related information mentioned above, the following criteria are considered suitable for defining the resource classification at Bisha for a pre-feasibility level:

• Material within the interpreted lithologic domain boundaries occurring within nominal 25 m sample spacing can be considered Indicated.
  

• Material lying within the lithologic domain boundaries and lying within 50 m of sampling within the domain boundaries is considered to be Inferred.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-4
	October 2004

• Material lying within the lithologic domain boundaries but not meeting these criteria is not classified.
  

• Material lying outside of the interpreted ore shell boundaries is not classified.
  

17.2.7 Resource Summaries

Table 17-1 summarizes the Indicated and Inferred Resources for the Oxide, Supergene, Primary Cu and Primary Zn Zones based on either Au, Cu or Zn cut-offs. Table 17-2 summarizes by Au cut-offs for the Oxide domain and by Cu for the Supergene domain. Table 17-3 summarizes the Primary domain based on a 2% Zn cut-off. A more complete summary for the mineral resources is provided in Section 17.9.4 of this report.

Table 17-1
Summary of the Bisha Mineral Resource Estimate (Brisebois, 2004)

Category	Zone	Cut-off	Tonnes (000’s)	Au g/t	Ag g/t	Cu %	Zn %
Indicated	Oxides	0.5g/t Au	4,984.1	6.51	30.0	0.10	0.08
	Supergene Cu	0.5% Cu	7,644.8	0.46	35.56	3.47	0.87
	Primary	2.0% Zn	1,711.5	0.74	29.59	0.97	3.07
	Primary Zn	2.0% Zn	8,413.3	0.76	58.27	1.12	9.04
	Total tonnes		22,753.7
Inferred	Oxides	0.5g/t Au	122.0	3.34	18.2	0.12	0.07
	Supergene Cu	0.5% Cu	185.6	0.09	30.14	3.26	1.04
	Primary	2.0% Zn	392.0	0.75	35.20	1.24	3.03
	Primary Zn	2.0% Zn	5,150.9	0.70	59.67	0.84	8.28
	Total tonnes		5,850.5

Table 17-2
Summary of the Oxide and Supergene Zone Resource Estimates (Brisebois, 2004)

Zone	Category	Domain	Cut-off Au g/t	Tonnes (000’s)	Au g/t	Ag g/t	Cu %	Zn %	Oz Au millions
Oxide	Indicated	Fe Oxide	0.5	3,653.4	6.65	20.32	0.10	0.08	0.781
Oxide	Indicated	Acidified	0.5	709.6	8.35	108.05	0.09	0.03	0.190
Oxide	Indicated	Breccia	0.5	621.1	3.62	9.39	0.09	0.07	0.072
Oxide	Indicated	Fe Oxide	1.0	3,469.5	6.97	20.79	0.10	0.08	0.778
Oxide	Indicated	Acidified	1.0	683.7	8.63	109.88	0.10	0.03	0.189
Oxide	Indicated	Breccia	1.0	519.1	4.19	10.65	0.09	0.07	0.069
Supergene	Indicated		Cut-off % Cu	Tonnes (000’s)	Au g/t	Ag g/t	Cu %	Zn %	lbs Cu millions
Supergene	Indicated		0.25	8,105.2	0.44	34.5	3.30	0.86	589
Supergene	Indicated		0.5	7,644.8	0.46	35.6	3.47	0.87	585
Supergene	Indicated		1.0	6,453.1	0.50	38.7	3.97	0.91	564

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-5
	October 2004

Table 17-3
Summary of the Primary and Primary Zn Resource Estimate (Brisebois, 2004)

Zone	Category	Cut-off Zn %	Tonnes (000’s)	Au g/t	Ag g/t	Cu %	Zn %	Contained Zn Millions lb	Contained Cu Millions lb
Primary	Indicated	2.0	1,711.5	0.74	29.6	0.97	3.07	115	37
Primary Zn	Indicated	2.0	8,413.3	0.76	58.3	1.12	9.04	1,680	207.7
Primary	Inferred	2.0	392.0	0.75	35.2	1.24	3.03
Primary Zn	Inferred	2.0	5,150.9	0.70	59.7	0.84	8.28

17.3

Geological Models and Creation of Vulcan Databases

17.3.1

Introduction

The geological modelling work was completed in Gemcom modelling software while grade estimation work was carried out using Vulcan modelling software. Section 17.3.2 discusses the loading of the geological models and assay databases into the Vulcan modelling system software. Section 17.3.3 discusses capping of high grades using the assay databases. Section 17.3.4 discusses compositing of the assays and the final EDA in support of estimation domains. Section 17.3.5 presents a summary of the mineralization controls used to domain the deposit for block grade estimation.

17.3.2

Geological Models and Assay Database

Nevsun and AMEC completed geological modelling. The geological model was developed using sectional spacing of 25 m for the basic interpretations that were subsequently rationalized in plan. Solid (wireframe) models were created for each principal mineralized or geological domain and became the basis for coding the block model.

The following lithologic codes were used to model the domains:

Breccia: REBX, HEBX, QZBX, SAPR

Oxide: FERC, FERU, CONG, HALF Acid: SOAP, SAND

Supergene: MSUL, SMSX, MDST

Primary: MSUL, SMSX, MDST

The near-surface resources were (Breccia, Oxide and Acid) were ultimately grouped as the Oxide Zone for resource reporting. The Primary domain was separated into Primary Cu and Primary Zn for estimation and reporting.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-6
	October 2004

The lithologic models were loaded to VULCAN using exported DXF files from GEMCOM. Various checks were carried out using facilities within VULCAN to successfully test the validity of the triangulations. Volumes were also compared to ensure a correct transfer.

The assay database was transferred directly from the MS ACCESS database and loaded to Vulcan using CSV files. The transfer was reviewed and validated in detail to ensure that the data was loaded correctly. No discrepancies were found.

17.3.3

Reduction of High Grades (Capping)

Grade capping is used in this resource model to limit the spatial extrapolation of the occasional, but generally isolated, anomalously high grades. For this study, AMEC has utilized the AMEC simulation-based approach²³. Choice of cap grades from probability plots was also used as a comparison. In the AMEC simulation method, the fundamental concept is that higher cap grades can be used where the density of data is higher and lower cap grades must be used where the density of data is lower. AMEC’s simulation method assumes that very high-grade assays are spatially independent, and that even small changes in the number of these present are directly translatable into significant changes in resource estimates. This would certainly be true if a nearest-neighbour estimator was used, and the effect could be more significant when kriging or another weighted average technique is used.

AMEC simulates the amount of high-grade mineralization which might be present by first “re-drilling” the deposit 1,000 times. In each case, the amount of metal attributable to a high-grade population is noted. The results are ordered, and the 20th percentile is chosen. If more samples are present, the 20th percentile of the simulated high-grade distribution will have a lower dispersion, and the 20th percentile will occur at a higher grade. This is termed the risk-adjusted metal. The risk-adjusted metal is added to the lower-grade material (assumed constant in the simulation process). The uncapped grade distribution is iteratively capped until it represents the sum of the risk-adjusted high-grade and the low-grade metal. The mine should exceed the risk-adjusted metal content in four years out of five. One year out of five it will do worse. The resources removed by the capping process cannot be reliably scheduled and should be considered Inferred because of its low confidence level. This material, therefore, should not be part of a Measured + Indicated Resource or a Proven + Probable Reserve. Over the life of the mine the metal removed from the resource estimate by capping should be recovered, but we cannot know when or how much. A disadvantage of this method is that it is deterministic; the cap grade may not necessarily coincide with a break in slope on a probability plot.

²³ Developed by MRDI, a predecessor of AMEC Mining & Metals.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-7
	October 2004

Given that the model is for pre-feasibility level purposes and that there are no available preliminary mine plans, the distributions used for the grade capping analysis consisted of the entire assay populations within the modelled lithologic domains. For feasibility level studies, it is recommended that capping analysis be carried out within the mineable area. The assay distributions used in the capping analysis are included in Appendix G.

AMEC’s simulation-based method used the following production rate assumptions provided by Nevsun:

• 2 Mtpy production rate in the Oxide and Supergene domains.
  

• 4 Mtpy production rate in the Primary domains.
  

Using these inputs and the declustered assay distribution, a percentage metal-at-risk was calculated for each metal to be estimated. The results are presented in Table 17-4.

Table 17-4
Metal-at-Risk Results from Capping Simulations

Metal	% Metal at Risk	Cap by Simulation	Cap by Probability Plot
Zn	2.54	22.5 %	28 %
Au	7.17	42.5 ppm	46 ppm
Ag	13.64	755 ppm	500 ppm
Cu	3.6	17.5 %	28 %
Pb	7.38	11 %	10 %

These values are typical for this style of deposit and at this stage of project development. The levels will likely decline somewhat as drilling continues on the Property. Ag is problematic but the very high grades are mainly restricted to the Acid domain, which is volumetrically contained and relatively small.

The cap grades suggested by the simulation method are included in Table 17-4 although these capping levels are not used directly (see implementation discussion below). For comparison purposes, cap grades chosen subjectively from the probability plots are also included in the table. In most case the comparison is quite good.

Since bulk density was not estimated (i.e. not estimated using the same approach as for the grades), capping analysis as applied here was not warranted. Bulk density determinations and the final values used in the resource model are discussed in Section 11.2.7.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-8
	October 2004

17.4

Compositing and Exploratory Data Analysis

17.4.1

Introduction

Exploratory data analysis (EDA) is the application of statistical tools to reveal characteristics of the data. In most cases, the results are used to assist in defining the ore controls or domains for block estimation. AMEC carried out data analysis on assays and composites. EDA included calculation of the mean and coefficient of variation, preparation of histograms, log-probability plots, scatterplots, and contact analysis.

Compositing was carried out using down-hole compositing routines in Vulcan. AMEC hand-checked a variety of composite calculations and found no errors in the calculated composites.

In the calculation of composites, AMEC used two methodologies. These included compositing within the logged geological codes and, secondly, compositing by starting and stopping on the boundaries of the solids modelled lithologic domains. Comparison of the composite distributions resulting from the two methods showed no significant differences. The composites that started and stopped on solids boundaries were accepted for use in estimation.

17.4.2

Choice of Composite Length

To evaluate composite length, AMEC studied the composites on section to assess the suitability of the composite lengths to the spatial trends seen in the mineralization as well as the scale of the lithologic domains. With the nominal spacing of drill sampling currently available in the deposit, the spatial trends observed in the data are on the order of ten’s of metres. Perhaps more importantly, resolution within the lithologic domains is probably best served with smaller composite lengths. It was concluded that the 3 m composite length is optimal for the deposit.

The compositing method does create smaller than 3 m composites where holes exit the modelled solids. Although not studied at this stage, AMEC recommends that composite length versus grade be assessed during the feasibility level studies.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-9
	October 2004

17.4.3

Statistical Distributions

Statistical analyses for the composites of the various metals are included in Appendix G. These include boxplots²⁴, histograms and probability plots for each metal by lithologic domain. Boxplots of the composite metal grades (Zn, Cu, Au, Ag, Pb) organized by lithologic domain coding are reproduced in Figures 17-1 to 17-5.

17.4.4

Contact Analysis

To further analyze the deposit, AMEC completed a thorough analysis of the behaviour of the metal grades adjacent to the major contacts within the lithologic model. Appendix G includes all of the contact analysis²⁵ results while the Table 17-5 gives a summary of the analyses in matrix form.

Figure 17-1
3 m Composite Statistical Comparison by Lithologic Domain - Zn (%)

²⁴ A boxplot is essentially the histogram of the data in a somewhat different format, which facilitates side by side comparison of populations. The box itself illustrates the 25th and 75th percentile and median, while the mean of the distribution is plotted as the dot. The minimum and maximum of the data are plotted at the end of the lines extending vertically above and below the box.

²⁵ This analysis plots the average grade of composites within bins of three-dimensional separation distances between composites identified as being on opposite sides of a given contact. The samples are identified based on the criteria posted at the top of each side of the contact plot. In each distance bin on the graph, the number of composites found within that distance from the contact is posted as a small number. The number of samples found in each group and their overall mean grade is also posted.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-10
	October 2004

Figure 17-2
3 m Composite Statistical Comparison by Lithologic Domain - Cu (%)

Figure 17-3
3 m Composite Statistical Comparison by Lithologic Domain - Au (g/t)

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-11
	October 2004

Figure 17-4
3 m Composite Statistical Comparison by Lithologic Domain - Ag (g/t)

Figure 17-5
3 m Composite statistical Comparison by Lithologic Domain - Pb (%)

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-12
	October 2004

Table 17-5
Summary of Contact Analysis Results

Domain	Zn	Cu	Au	Ag	Pb
Breccia - Oxide	Soft	Soft	Soft to Moderate	Soft to Moderate	Soft
Oxide - Acid	Soft	Soft	Very Hard	Very Hard	Moderate
Acid – Supergene	Soft	Hard	Very Hard	Moderate	Discontinuous at < 6m (in acid)
Supergene - Primary Cu	Hard	Moderate to gradational	Soft	Soft	Soft
Supergene - Primary Zn	Hard	Moderate to gradational	Soft	Soft	Soft
Primary Cu - Primary Zn	Hard	Soft	Soft	Soft	Soft

The estimations domains were chosen for each metal by combining the results of the contact analysis detailed above, the statistical analyses, preliminary variography and geological considerations. These domains are summarized below.

17.4.5

Summary of Ore Controls

The modelled domains provide a reasonable global framework to limit the estimation and extrapolation of economic gold mineralization within the deposit. Exploratory data analysis of assays and composites within and outside the domain boundaries reveal several pertinent issues:

• There are different trends in the various metals with respect to the modelled lithologic domains. To both a lesser and greater degree, each metal exhibits unique trends within the deposit.
  

• The distributions are generally well-behaved with relatively low coefficients of variation.
  

• Grade profiles for each metal across the various domain boundaries illustrate a variety of behaviours ranging from soft (little change in average grade) to hard boundaries (abrupt change in average grade very close to the contact).
  

• Analyses of correlations between the various metals show complex relationships.
  

To address these issues, the following concepts were developed to assist in domaining the deposit for grade estimation:

• The outer boundaries of the lithologic domains were used to form hard boundaries to separate the potentially economic mineralized material from the largely un-mineralized or sub-economic material outside of the interpreted domains.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-13
	October 2004

• There was no estimation of grade outside of the lithologic domains.
  
• Within the lithologic domains, and groupings of these domains identified during the EDA phase of the work, ordinary kriging was determined to be a suitable estimator for the pre-feasibility results.
  
• Correlations between various metals, while complex, should be further studied in the resource modelling for feasibility level studies. The current model makes a reasonable effort to reproduce the correlations between the various metals but could likely be improved. An improvement in this area would be unlikely to affect the global resource inventory but may improve local estimation in the model.
  

The domains used to estimate each metal are summarized in Table 17-6.

Table 17-6
Estimation Domains

Domain	Zn	Cu	Au	Ag	Pb
1	Bx+Ox+Acid+Supg	Bx+Ox+Acid	Bx+Ox	Bx+Ox	Bx+Ox+Acid
2	Primary Cu	Supg+Primary(all)	Acid	Acid	Supg+Primary(all)
3	Primary Zn		Supg+Primary(all)	Supg+Primary(all)

17.5

Variography

17.5.1

Introduction

For most mineral deposits, the measure of spatial variability for a given metal depends on both the separation distance between points of measurement and the direction from position to position. Variability increases and correlation decreases with the sample to sample separation distance. When the rate of change in variability is dependant on the direction, the measure of spatial variability is described as anisotropic. Usually this anisotropy can be characterized in terms of an ellipsoid, which has axes of anisotropy. There are several functions that measure variability; the most frequently used is the semi-variogram. The nugget effect of a variogram model indicates short-range variability, which is often difficult to separate from the variability introduced in sample collection, preparation and assaying. Variograms are often modelled using a composite function, each component of which may have its own sill and range (point at which the contribution to variability becomes constant).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-14
	October 2004

The spatial variability for the various metals in the ore zones was defined using a correlograms²⁶ function. The correlogram was probably the first measure of continuity (the converse of variability) developed. It measures the correlation coefficient between two sets of data, comprising values at the heads and values at the tails of vectors with similar direction and magnitude. Srivastava and Parker (1988) found that the correlogram provided a stable estimate of spatial continuity. It is most applicable where the variance is defined and is directionally stationary. For ease of modelling, the correlogram value is subtracted from one and is presented in a similar graphical form as the variogram. In this report the correlograms presented this way are referred to as variograms.

At Bisha, spatial variability is complicated by local trends of high and low-grade mineralization. Some of the trends are naturally occurring. However, some of these trends emerge from the necessity of combining populations where there is not enough data to support two sub-populations. Some of the poor correlation is caused, for example, by the isolation of the Acid domain. This is more a practical necessity rather than a geostatistical decision due to its anomalous nature in a comparatively small volume. All of these factors have led to variograms that, in a few cases, exhibit relatively poor structure. In most cases, variograms are moderate to good leading to relatively confident estimation results.

17.5.2

Variogram Models

Variograms were calculated for the five metals within each of the estimation domains discussed above. The variogram modelling was completed within the SAGE 2000 variogram software using the 3 m composite database.

Appendix G contains output (digital format) from SAGE 2000 for the various domains and models. Table 17-7 summarizes the variogram models for the various metals and estimation domains. In Table 17-7, each metal is listed along with its estimation domains. Within each column the parameters that describe the variogram model are tabulated. Each model is a two structure spherical correlogram. The description of the ellipsoid describing each structure has been entered using conventional azimuth and dip (positive upward) measurement to describe the orientation of each of the axes. The ranges are entered in metres.

²⁶ Variograms and correlograms are both functions of the vectorial oriented distance measuring the spatial correlation or continuity of the RF (random function) Z under study. One minus the correlogram is AMEC’s common tool, which gives an estimate of the variogram with a unit sill. Definitions and notations:

variogram: ?(h) = ½ E [ ( Z(x) - Z(x+h)) 2 ]

correlogram : r(h) = [E( Z(x) . Z(x+h)) - E(Z(x)E(Z(x+h)]/(sx . sx+h) = 1-g? (h)/ s 2 with E [ f(Z(x)) ]

meaning the mathematical expectation of a function f applied on RF Z for all locations x over the study domain D, gx standing for the standard deviation of Z on the domain Dx of points which can be used as first points(.x) in pairs (.x ,.x+h) at a distance h.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-15
	October 2004

In general, the Breccia, Oxide, Acid and Supergene variograms exhibit increased correlation in a horizontal orientation with limited anisotropy within that plane.

In some cases, the orientation of the second structure on the variogram models tends to follow the overall attitude of the deposit structure (the general north-south strike). This is especially true in the base metal variograms in the Primary domains. The first structure of the variograms are typically aligned on something other than the general trend and are likely reflecting an averaging of local shorter scale trends with high correlation.

The Au and Ag variograms are generally high nugget variograms exhibiting better correlation in more flat-lying orientations. The variograms have poor to moderate structure. In the Breccia and Oxide domains, the down-hole variograms are surprisingly good.

Table 17-7
Variogram Models

	Zn			Cu		Au			Ag			Pb
	Prim.Zn	Prim.Cu	brx + ox + acid + supg	brx + ox + acid	Supg + Cu + Zn	Brx + Ox	Acid	Supg + Cu + Zn	Brx + Ox	Acid	Supg + Cu + Zn	brx + ox + acid	Supg + Cu + Zn
Nugget (C0)	0.250	0.200	0.050	0.100	0.150	0.080	0.575	0.450	0.100	0.600	0.450	0.400	0.550
Sill Structure1(C1)	0.599	0.623	0.900	0.616	0.483	0.644	0.102	0.442	0.794	0.001	0.312	0.479	0.227
Sill Structure2(C2)	0.151	0.177	0.050	0.284	0.367	0.276	0.324	0.108	0.106	0.399	0.238	0.121	0.223
Z1 Azimuth (Minor)	86	151	70	72	61	84	230	29	61	78	80	100	90
Z1 Dip (Minor)	57	80	56	39	32	55	56	74	32	4	58	47	51
Y1 Azimuth (Semi-major)	6	47	349	48	343	356	113	29	357	166	3	13	349
Y1 Dip (Semi-major)	-6	3	-6	-48	-19	-1	17	-16	-35	-32	-8	-3	9
X1 Azimuth (Major)	100	136	83	152	99	86	193	119	121	174	97	106	72
X1 Dip (Major)	-32	-10	-33	-12	-51	-35	-29	0	-39	58	-31	-43	-37
Str1 : Range along Z'	10.1	37.2	15.8	20.2	11.7	29.0	5.5	12.1	29.8	20.7	12.2	24.6	21.1
Str1 : Range along Y'	69.5	43.8	100.3	17.9	106.9	13.2	71.0	49.5	46.6	18.9	32.7	58.6	76.8
Str1 : Range along X'	8.4	10.0	14.2	1.2	17.9	6.4	95.7	4.9	7.5	19.0	7.6	11.5	11.2
Z2 Azimuth (Minor)	38	116	338	235	264	224	157	112	79	240	164	276	141
Z2 Dip (Minor)	70	37	11	19	4	20	19	61	-7	27	39	33	49
Y2 Azimuth (Semi-major)	346	20	68	334	354	312	67	5	347	317	28	4	4
Y2 Dip (Semi-major)	-13	7	2	25	0	-4	1	9	-13	-24	41	-3	32
X2 Azimuth (Major)	80	100	166	113	89	30	154	90	195	11	95	89	79
X2 Dip (Major)	-15	-52	79	58	86	70	-71	-27	-75	52	-24	56	-22
Str 2 : Range along Z'	334.5	98.2	491.3	110.5	84.6	63.6	322.9	108.9	123.7	15.4	150.5	102.3	303.2
Str 2 : Range along Y'	225.9	308.3	1048.4	66.1	856.1	270.2	190.1	640.0	343.6	134.3	699.4	229.6	350.7
Str 2 : Range along X'	28.3	20.9	213.2	27.3	43.5	26.6	14.6	41.4	58.7	48.2	69.4	34.3	38.8

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-16
	October 2004

17.6

Estimation Plan

17.6.1

Block Model Setup

The block model in the Vulcan® modelling system was set up with the dimensions and parameters shown in Table 17-8. The size of the blocks was selected by visible inspection of the geometry of the lithologic zones as well as the local trends observed in the various metal grades. This had led to a large number of blocks, however, the short vertical height of the blocks is though necessary to help delineate the ACID domain.

Table 17-8
Block Model Parameters

Parameters	Easting	Northing	Elevation	Bearing	Dip
Origin	339100	1715000	0	90	0
	x	y	z
Dimensions (m)	5.0	5.0	3.0

Figure 17-6 shows two views of the block model orientation with the interpreted lithologic domains included to assist in visualizing the deposit orientation with respect to the block model (in the figure, X is pointing east, Y is pointing north, z is pointing upwards).

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-17
	October 2004

Figure 17-6
Orientation Views of the Block Model

17.6.2

Estimation Plan Parameters

The estimation plan includes the following items:

• Storage of predominant domain code, percentage mineralization and percentage of each individual domain in the block.
  

• Application of density in each block based on the predominant lithologic domain in the block.
  

• Estimation of metal grades by ordinary kriging using two passes – one pass for blocks close to sample data and a second pass for blocks further from drill sampling.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-18
	October 2004

17.6.3

Grade Estimation

Table 17-9 entries summarize the search ellipse dimensions for the estimation passes in the various estimation domains. All estimations used a minimum of three composites to estimate a block and a maximum of twelve. A maximum of two composites from a given drill hole was enforced.

Table 17-9
Search Ellipse Parameters

	Search Ellipse
	Major	Semi-Major	Minor	Azimuth	Dip
Pass 1 (Bx+Ox+Acid+Supg)	50	50	15	350	0
Pass 2 (Bx+Ox+Acid+Supg)	200	200	15	350	0
Pass 1 (Primary)	50	50	15	350	71
Pass 2 (Primary)	200	200	15	350	71
Note: Azimuth is the direction of the rotated x axis (Major).
Note: Dip is the positive down rotation of the y' axis around x' (Semi-Major)

17.6.4

Implementation of Grade Capping Strategy

In order to apply the reduction in metal-at-risk discussed in section 17.3.3, the estimation plan was re-run in a new block model with the addition of parameters in the estimation to restrict high grade composites during the block estimation. The cap grades were adjusted by trial and error to arrive at the desired reduction in metal.

An improved methodology, which is recommended for models to be used for the feasibility level studies, would be to use the facilities within VULCAN which allow the restriction of high grade samples based on a distance threshold from the block centroids. The advantage to this approach is that high grade composites (uncapped) may be used to estimate blocks when the composites are close enough to a given block.

17.6.5

Discussion

AMEC recommends that the use of oriented search ellipses for different areas of the deposit be attempted for upcoming resource models. This approach is useful to improve the ability of the estimation to efficiently select samples in areas of the deposit where the geometry of the ore body departs from the average orientation.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-19
	October 2004

17.7

Bulk Density

Section 11.2.7 provides a summary of AMEC’s review of the bulk density information. Table 17-10 provides the average values and the rock types of the individual measurements used to determine the bulk density values for each domain.

Tables 11-6 and 17-10 summarize the number of values, range of values and the average bulk density values used for each domain in the resource model. AMEC prepared descriptive statistics of bulk density for each rock type that comprised the domain. Each domain has a primary style of mineralization but other lithologies are also captured in the construction of 3D wireframes. The percentage of volume of those other rock types was determined for each domain to ensure there was not a bias in the bulk density collection that would favour a lower or higher bulk density value. The average bulk density values were calculated for each of the domains and were determined to be acceptable for the resource estimation.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-20
	October 2004

Table 17-10
Bulk Density Values by Domain and Rock Type within Each Domain

SG	Zone	Count	Min	Max	St. Dev.	Mode	Median	SG	Rock Code	Count	Min	Max	St. Dev.	Mode	Median
2.21	BX	15	1.77	3.19	0.35	2.01	2.13	2.44	CONG	1	-	-	-	-	-
								2.28	HALF	1	-	-	-	-	-
								2.55	HEBX	4	2.03	3.19	0.51	N/A	2.50
								2.11	QZBX	4	2.01	2.15	0.07	N/A	2.14
								1.95	SAPR	5	1.77	2.13	0.13	N/A	1.93
3.32	OXID	36	1.89	4.05	0.53	3.17	3.43	3.46	FERC	13	3.00	4.05	0.31	3.17	3.53
								3.43	FERU	17	1.89	3.95	0.50	3.7	3.55
								2.15	HALF	1	-	-	-	-	-
								3.02	HEBX	3	2.21	3.57	0.72	N/A	3.29
								2.48	QZBX	2	2.31	2.65	0.24	N/A	2.48
2.08	ACID	6	1.10	2.82	0.66	#N/A	2.28	2.08	SOAP	6	1.10	2.82	0.66	N/A	2.28
4.15	SUPG	115	1.38	5.05	0.80	4.17	4.39	2.41	FPDK	2	2.32	2.51	0.13	N/A	2.41
								1.89	MDST	2	1.67	2.11	0.31	N/A	1.89
								4.25	MSUL	109	1.38	5.05	0.68	4.17	4.43
								3.90	SMSX	1	-	-	-	-	-
								1.62	SOAP	1	-	-	-	-	-
4.39	PRIM	176	2.32	4.99	0.56	4.78	4.56	2.40	FPDK	2	2.40	2.40	0.00	2.40	2.40
								2.81	MAFD	1	-	-	-	-	-
								2.76	MAFT	1	-	-	-	-	-
								3.51	MDST	3	2.78	4.59	0.95	N/A	3.16
								4.58	MSUL	146	3.49	4.99	0.31	4.78	4.65
								3.78	SMSX	19	3.29	4.66	0.36	3.29	3.80
								2.32	SOAP	1	-	-	-	-	-
								3.13	STSX	3	3.05	3.26	0.11	N/A	3.09
4.43	PRIM_ZN	260	2.00	5.39	0.49	4.80	4.56	4.09	MAFT	1	-	-	-	-	-
								3.34	MDST	8	2.29	4.80	0.89	N/A	2.99
								4.51	MSUL	240	2.00	5.39	0.36	4.80	4.59
								2.75	QZBX	1	-	-	-	-	-
								3.55	SMSX	6	2.75	4.82	0.65	N/A	3.23
								3.42	STSX	3	2.87	4.46	0.90	N/A	2.94
								3.92	XTUF	1	-	-	-	-	-

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-21
	October 2004

17.8

Model Validation

17.8.1

Introduction

To validate the kriged estimates a number of checks were carried out. These included:

• Visual inspection of kriged results on plans and sections.
  
• Comparison of kriged and nearest neighbour statistics.
  
• A change of support check.
  

17.8.2

Visual Inspection

AMEC studied the block estimates by visual inspection of plans and sections against a background of drill hole composite data. AMEC found that estimates appeared to honour the composites reasonably well. The check was generally restricted to the Indicated and Inferred Resource blocks in the model.

17.8.3

Comparison to Nearest Neighbour Estimation

AMEC calculated nearest neighbour estimates (polygonal assignment from 3 m composites) for each of the metals. The nearest neighbour estimates were developed using the estimation plan created for the kriged estimation.

Table 17-11 summarizes the comparisons between nearest neighbour and kriged results. The results are calculated at a zero cut-off for those blocks in the Indicated category. The differences are very close for most metals and are well within accepted norms. Note that these comparisons are done on an un-capped basis.

Table 17-11
Comparison of Kriged Versus Nearest Neighbour – Indicated Blocks

Bisha : Indicated Resource Blocks
(at a 0.0 cut-off)	Kriged Estimate	Nearest Neighbour	Difference (%)
Ag	42.09	42.37	-0.66
Zn	2.84	2.85	-0.31
Au	2.19	2.19	-0.12
Pb	0.28	0.29	-3.96
Cu	1.47	1.45	1.24

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-22
	October 2004

Figures 17-7 to 17-9 show profiles through the Bisha Deposit, which compare the kriged estimate to the nearest neighbour estimate. In general the agreement is very good. The trend of the average kriged block grade follows the average of the nearest neighbour grade quite closely. There are some areas where the two profiles appear to diverge significantly but in these cases the accompanying tonnage profile shows that the amount of estimated material in these areas drops off considerably.

17.8.4

Change of Support Check

One of the objectives of the kriging plan developed for the resource model is to tune the variability of individual kriged block grades such that they would have a similar grade tonnage distribution to the selective mining unit (SMU) on which final selection is to be made.

The coefficient of variation (CV) is used when comparing the population statistics of block estimates to the theoretical SMU distribution. In the case of Bisha, several of the estimation domains were checked. Given the pre-feasibility level of the project, several SMU sizes were analyzed. The principal analyses were done using the Indicated blocks. Similar analyses were carried out for all blocks for comparative purposes. Table 17-12 summarizes some of the more pertinent results.

Table 17-12
Change of Support Analysis

	Indicated Blocks
	Composite Support		SMU Normalized	Target CV	Estimated Blocks
Zone	Mean	CV	Blk Disp. Var.		CV	%diff	Notes
Zn, Cu+Zn Domain	4.59	1.320	0.6576	1.0704	1.0200	4.7	6hx5x5  SMU informed by 5x5m prod. samples
Au, Bx+Ox Domains	6.16	1.700	0.6709	1.3924	0.9800	29.6	6hx5x5  SMU informed by 5x5m prod. samples
Au, Bx+Ox Domains	6.16	1.700	0.5586	1.2706	0.9800	22.9	6hx5x10  SMU informed by 5x5m prod. samples
Au, Bx+Ox Domains	6.16	1.700	0.4323	1.1177	0.9800	12.3	6hx10x15  SMU informed by 5x5m prod. samples
Au, Acid Domains	8.55	2.150	0.4007	1.3610	0.9800	28.0	6hx5x5  SMU informed by 5x5m prod. samples
Cu, Supg+Cu+Zn Domains	1.66	1.730	0.6898	1.4368	1.2700	11.6	6hx5x5  SMU informed by 5x5m prod. samples

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-23
	October 2004

Figure 17-7
Grade Profiles Comparing Kriged to Nearest Neighbour Block Estimates in the Indicated Blocks (Au and Estimated Tonnage)

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-24
	October 2004

Figure 17-8
Grade Profiles Comparing Kriged to Nearest Neighbour Block Estimates in the Indicated Blocks (Ag and Zn)

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-25
	October 2004

Figure 17-9
Grade Profiles Comparing Kriged to Nearest Neighbour Block Estimates in the Indicated Blocks (Pb and Cu)

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-26
	October 2004

In Table 17-12, target CV’s derived from the variogram for the metal and estimation domain and the chosen SMU size are shown beside the CV’s of the final estimated blocks. The difference between the desired CV and that achieved by the estimation is shown in the ‘%diff’ column.

For Zn in the Primary domain, the current model is showing a difference of 4.7% on the CV for the estimated blocks versus that predicted for the SMU size. This indicates that the Zn estimates are reasonably well tuned for 6hx5x5m SMU selectivity. In fact, in the case of Zn, the blocks estimates can be relatively easily adjusted to a different selectivity if desired in future resource models. The Zn distribution is relatively well behaved and lends itself well to Ordinary Kriging for block estimates.

In the case of Au in the Breccia and Oxide domain, the distributions are not as easily handled. Three SMU sizes were used to help demonstrate the degree of selectivity currently supported by estimation. The results show that the Au estimates are somewhat smoothed in this domain. These areas, however, should be extracted under open pit mining so the large SMU size will be used. At a 6hx10x15m SMU size, the estimates in the current model are quite close to the desired SMU distribution. It is likely that a refinement in the Au estimation technique will be necessary for these domains upcoming resource models. The refinement should be directed at improving the selectivity of the block estimates. It is also likely that more mining decisions will have been made based on the follow-on work from this model. There are several possibilities for refinement including, the further sub-division of these two domains, changes in the estimation plan and/or changes in the estimation technique. A refinement such as this is unlikely to change the global resource base (metal content), however, it may locally change grades and tonnages in the blocks as the desired selectivity is achieved.

The Cu estimation in the Supergene and Primary domains was also reviewed and found to be reasonable although some improvement may be possible in upcoming resource models. One change that should be explored would be the further sub-division of this domain in two or more sub-domains.

During the future resource modelling, AMEC recommends that further change of support analyses should be completed. This would consist of a Hermitian change of support analysis of the Nearest Neighbour distribution (declustered composite distributions) in order to assess and refine the tonnes and grade near the cut-offs of interest for the various domains.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-27
	October 2004

17.9

Resource Classification and Summaries

17.9.1

Introduction

The mineral resources have been classified according to guidelines and logic summarized within the CIM Definitions referred to in National Instrument 43-101. Resources were classified as Indicated or Inferred by considering geology, sampling and grade estimation aspects of the model. For geology, consideration was given to the confidence in the interpretation of the lithologic domains. For sampling, consideration was given to the number and spacing of composites, the orientation of drilling and the reliability of sampling. For estimation, consideration was given to how well the kriged grades reflect the composites and how high-grade composites were projected into plan and section. For the estimation results, consideration was given to the confidence with which grades were estimated.

17.9.2

Confidence Intervals for Grade Estimation

The following approach is used by AMEC for assisting in the classifying blocks into appropriate resource classes. The methodology utilizes confidence limits analysis as a starting point, yielding theoretical justification for initial choices of sample spacing and, if desired, estimation variance thresholds.

AMEC has generally found that for base and precious metals deposits, sampling must be sufficient to estimate the tonnage, grade and metal content on annual production increments ± 15% at 90% confidence in order to define an Indicated Resource. That is, when the stated confidence limits are met, cross-sectional and level plan interpretations show continuity with respect to orebody outlines and grade. In addition AMEC has often found that annual cash flow projections can accommodate a 15% drop in tonnage, grade or metal content without severely affecting project viability. Also, many projects are designed and operated in such a way that a 15% shortfall can be made-up by rescheduling production. More severe shortfalls are difficult to overcome. Finally, the planning horizon for feasibility studies is generally annual increments, and a ± 15% error level is often applied to capital and operating costs when conducting a feasibility study. Hence a balanced approach requires a similar degree of confidence in the resources/reserves.

For development of confidence intervals, idealized blocks approximating the production from one month are estimated by ordinary kriging using different grids of samples. For a mine with many working faces, the assumption is made for practical purposes that the one month production comes from one large block. For the case of Bisha, the current production rate projections are 2 Mtpy in Oxides and Supergene and 4 Empty in Primary material. The dimensions of a large ideal block representing the volume are calculated and

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-28
	October 2004

the correlograms for grade are used in kriging the ideal block. The large blocks are analyzed using theoretical drill holes at different pattern spacings.

The kriging variances from these ideal blocks are first converted to relative variances. In the case of the correlogram, converting to a relative variance involves multiplying by the coefficient of variation or alternatively rescaling the sill of the correlogram by the relative variance. The large block kriging variance is then divided by twelve (assuming approximate independence in the production from month to month) to scale to a variance for yearly ore output. The square root of this kriging variance is then used to construct confidence limits under the assumption of normally distributed errors of estimation. For example, if the kriging variance for a block is g2m, then the kriging variance for the year is g2y = g2m/12. The 90% confidence limits are then C.L. = ±1.645 x gy.

Table 17-13 shows the 90% confidence interval results for Bisha metal estimations for several of the domains of interest.

17.9.3

Discussion

Based on the results of confidence interval calculation for the block grade estimations, the following points relate to the resource classification criteria:

• Based on the various results, confidence interval calculations would suggest that 25 to 30 m spaced sampling is sufficient to define an Indicated Resource.
  

• In some cases, such as Cu in the Supergene and Primary domain, the ability to estimate the Cu grade with reasonable confidence is more than 25m spacing.
  

• In the case of Au in the Acid, confidence in estimation is quite low.
  

Synthesizing the above notes as well as other related information, the following criteria are considered suitable for defining the resource classification at Bisha for a pre-feasibility level of study:

• Material within the interpreted lithologic domain boundaries occurring within nominal 25 m sample spacing can be considered Indicated.
  

• Material lying within the lithologic domain boundaries and lying within 50 m of sampling within the domain boundaries is considered to be Inferred.
  

• Material lying within the lithologic domain boundaries but not meeting these criteria is not classified.
  

• Material lying outside of the interpreted ore shell boundaries is not classified.
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-29
	October 2004

Table 17-13
90% Confidence Intervals for Metal Estimation in Various Domains

Estimation Domain	Spacing (m)	CI(year)	CI(quarter)	CI(month)
Zn, Cu+Zn Domain	15	+/-5.9%	+/-11.8%	+/-20.4%
	20	+/-8.1%	+/-16.3%	+/-28.2%
	25	+/-9.7%	+/-19.3%	+/-33.5%
	30	+/-10.8%	+/-21.6%	+/-37.5%
	35	+/-11.7%	+/-23.5%	+/-40.6%
	40	+/-12.8%	+/-25.7%	+/-44.5%
Zn, Bx+Ox+Acid+Supg Domain	15	+/-7.4%	+/-14.9%	+/-25.8%
	20	+/-10.7%	+/-21.3%	+/-36.9%
	25	+/-12.4%	+/-24.8%	+/-42.9%
	30	+/-13.3%	+/-26.6%	+/-46.1%
	35	+/-13.9%	+/-27.8%	+/-48.1%
	40	+/-14.9%	+/-29.7%	+/-51.5%
Au, Bx+Ox Domains	15	+/-9.6%	+/-19.1%	+/-33.1%
	20	+/-13.9%	+/-27.9%	+/-48.3%
	25	+/-16.2%	+/-32.5%	+/-56.2%
	30	+/-18.2%	+/-36.3%	+/-62.9%
	35	+/-19.3%	+/-38.7%	+/-67.0%
	40	+/-23.7%	+/-47.5%	+/-82.2%
Au, Acid Domains	15	+/-13.1%	+/-26.3%	+/-45.5%
	20	+/-17.2%	+/-34.3%	+/-59.5%
	25	+/-19.2%	+/-38.5%	+/-66.6%
	30	+/-21.1%	+/-42.3%	+/-73.2%
	35	+/-23.2%	+/-46.3%	+/-80.3%
	40	+/-25.5%	+/-51.0%	+/-88.4%
Au, Supg+Cu+Zn Domains	15	+/-9.3%	+/-18.7%	+/-32.3%
	20	+/-12.2%	+/-24.4%	+/-42.3%
	25	+/-13.8%	+/-27.5%	+/-47.7%
	30	+/-14.9%	+/-29.8%	+/-51.6%
	35	+/-15.5%	+/-30.9%	+/-53.5%
	40	+/-16.3%	+/-32.6%	+/-56.5%
Cu, Bx+Ox+Acid Domains	15	+/-11.0%	+/-21.9%	+/-38.0%
	20	+/-12.3%	+/-24.6%	+/-42.6%
	25	+/-14.0%	+/-28.0%	+/-48.5%
	30	+/-14.9%	+/-29.9%	+/-51.8%
	35	+/-15.6%	+/-31.2%	+/-54.0%
	40	+/-17.8%	+/-35.6%	+/-61.6%
Cu, Supg+Cu+Zn Domains	15	+/-7.7%	+/-15.3%	+/-26.5%
	20	+/-10.6%	+/-21.3%	+/-36.8%
	25	+/-12.3%	+/-24.7%	+/-42.7%
	30	+/-13.5%	+/-26.9%	+/-46.6%
	35	+/-14.3%	+/-28.5%	+/-49.4%
	40	+/-15.7%	+/-31.4%	+/-54.3%

The criteria are implemented using stored distances to closest composite and closest two composites. In particular, to delineate the Indicated material some latitude is allowed whereby the nominal 25 m grid can be defined by two holes at this spacing. This approach is necessary given the geometry of the deposit.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-30
	October 2004

17.9.4

Resource Summaries

Table 17-14 summarizes the Indicated and Inferred Resources for all domains based on Zn cut-offs. Table 17-15 summarizes by Au cut-offs for the near-surface domains and Table 17-16 summarizes the Cu resource in the Supergene domain.

Table 17-14
Indicated and Inferred Resources by Zn Cut-offs

Indicated Resources (by Zn Cut-offs)						Inferred Resources (by Zn Cut-offs)
Domain	Tonnes (1000's)	Zn (%)	Cu (%)	Au (g/t)	Ag (ppm)	Domain	Tonnes (1000's)	Zn (%)	Cu (%)	Au (g/t)	Ag (ppm)
2.0% Zn Cut-off						2.0% Zn Cut-off
Brx+Ox+Acid	0.0	0.00	0.00	0.00	0.0	Brx+Ox+Acid	0.0	0.00	0.00	0.00	0.0
Supergene	438.3	4.47	3.60	0.75	54.3	Supergene	0.3	2.21	3.12	0.73	21.9
Primary Cu	1,711.5	3.07	0.97	0.74	29.6	Primary Cu	392.0	3.03	1.24	0.75	35.2
Primary Zn	8,413.3	9.04	1.12	0.76	58.3	Primary Zn	5,150.9	8.28	0.84	0.70	59.7
Total	10,562.1	7.88	1.20	0.76	53.5	Total	5,543.1	7.91	0.87	0.70	57.9
4.0% Zn Cut-off						4.0% Zn Cut-off
Brx+Ox+Acid	0.0	0.00	0.00	0.00	0.0	Brx+Ox+Acid	0.0	0.00	0.00	0.00	0.0
Supergene	175.4	7.02	3.37	0.78	60.9	Supergene	0.0	0.00	0.00	0.00	0.0
Primary Cu	286.9	5.66	1.03	0.89	34.1	Primary Cu	65.3	4.76	1.58	0.83	34.3
Primary Zn	7,789.4	9.51	1.14	0.77	60.0	Primary Zn	4,242.9	9.40	0.90	0.71	63.8
Total	8,251.4	9.32	1.18	0.78	59.1	Total	4,308.1	9.33	0.91	0.71	63.4
6.0% Zn Cut-off						6.0% Zn Cut-off
Brx+Ox+Acid	0.0	0.00	0.00	0.00	0.0	Brx+Ox+Acid	0.0	0.00	0.00	0.00	0.0
Supergene	87.2	9.24	3.19	0.79	65.6	Supergene	0.0	0.00	0.00	0.00	0.0
Primary Cu	78.8	7.84	0.89	0.92	37.3	Primary Cu	4.8	8.05	1.52	0.78	53.7
Primary Zn	6,422.2	10.54	1.16	0.78	63.3	Primary Zn	3,161.1	10.95	0.99	0.74	71.3
Total	6,588.1	10.39	1.18	0.78	63.0	Total	3,165.9	10.95	0.99	0.74	71.2

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-31
	October 2004

Table 17-15
Indicated and Inferred Resources by Au Cut-offs

Indicated Resources (by Au Cut-offs)						Inferred Resources (by Au Cut-offs)
Domain	Tonnes (1000's)	Au (g/t)	Zn (%)	Cu (%)	Ag (ppm)	Domain	Tonnes (1000's)	Au (g/t)	Zn (%)	Cu (%)	Ag (ppm)
0.5g/t Au Cut-off						0.5g/t Au Cut-off
Bx	621.1	3.62	0.07	0.09	9.39	Bx	2.4	3.66	0.19	0.19	5.87
Ox	3,653.4	6.65	0.08	0.10	20.32	Ox	59.5	4.80	0.10	0.12	14.80
Acid	709.6	8.35	0.03	0.09	108.05	Acid	19.1	5.53	0.02	0.24	67.16
Total	4,984.1	6.51	0.07	0.10	31.45	Total	81.0	4.94	0.08	0.15	26.88
1.0g/t Au Cut-off						1.0g/t Au Cut-off
Bx	519.1	4.19	0.07	0.09	10.65	Bx	2.2	3.92	0.17	0.20	6.33
Ox	3,469.5	6.97	0.08	0.10	20.79	Ox	54.4	5.17	0.10	0.13	15.82
Acid	683.7	8.63	0.03	0.10	109.88	Acid	19.1	5.53	0.02	0.24	67.16
Total	4,672.3	6.90	0.07	0.10	32.70	Total	75.7	5.22	0.08	0.16	28.50
1.5g/t Au Cut-off						1.5g/t Au Cut-off
Bx	425.5	4.84	0.06	0.10	12.02	Bx	1.7	4.71	0.14	0.22	7.15
Ox	3,289.4	7.28	0.08	0.10	21.44	Ox	50.8	5.45	0.11	0.13	16.76
Acid	660.5	8.89	0.03	0.09	111.77	Acid	18.3	5.72	0.02	0.25	69.04
Total	4,375.4	7.29	0.07	0.10	34.16	Total	70.8	5.50	0.09	0.16	30.04
2.0g/t Au Cut-off						2.0g/t Au Cut-off
Bx	357.5	5.43	0.06	0.10	13.22	Bx	1.5	5.13	0.16	0.25	7.69
Ox	3,079.7	7.66	0.08	0.10	22.21	Ox	45.9	5.85	0.11	0.14	18.26
Acid	634.1	9.19	0.03	0.09	113.47	Acid	17.2	5.96	0.02	0.07	64.33
Total	4,071.3	7.70	0.07	0.10	35.63	Total	64.6	5.86	0.09	0.12	30.28

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-32
	October 2004

Table 17-16
Indicated and Inferred Resources by Cu Cut-offs

Indicated Resources (by Cu Cut-offs)						Inferred Resources (by Cu Cut-offs)
	Tonnes (1000's)	Cu (%)	Au (g/t)	Zn (%)	Ag (ppm)		Tonnes (1000's)	Cu (%)	Au (g/t)	Zn (%)	Ag (ppm)
0.25% Cu Cut-off						0.25% Cu Cut-off
Supg	8,105.2	3.30	0.44	0.86	34.52	Supg	208.2	2.95	0.09	1.03	29.12
0.50% Cu Cut-off						0.50% Cu Cut-off
Supg	7,644.8	3.47	0.46	0.87	35.56	Supg	185.6	3.26	0.09	1.04	30.14
0.75% Cu Cut-off						0.75% Cu Cut-off
Supg	7,048.5	3.71	0.48	0.90	37.06	Supg	150.4	3.87	0.08	1.06	31.79
1.0% Cu Cut-off						1.0% Cu Cut-off
Supg	6,453.1	3.97	0.50	0.91	38.69	Supg	141.7	4.06	0.08	1.07	32.26

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 17-33
	October 2004

18.0

OTHER RELEVANT DATA AND INFORMATION

No other relevant data or information has been provided to AMEC that should be included in this report.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 18-1
	October 2004

19.0

REQUIREMENTS FOR TECHNICAL REPORTS ON PRODUCTION AND DEVELOPMENT PROPERTIES

This section is not applicable.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 19-1
	October 2004

20.0

CONCLUSIONS AND RECOMMENDATIONS

20.1

Conclusions

The Bisha Main Zone deposit is a large precious metal rich (Au, Ag), base metal rich (Cu, Zn, Pb) volcanogenic massive sulphide deposit hosted within a bimodal sequence of weakly stratified, predominantly tuffaceous metavolcanic rocks of the Nacfa Terrane greenstone belt. Oxidation of the near-surface of the deposit has created a zonation of four principal zones of mineralization within the Bisha Main Zone including: (1) a near-surface gold rich oxide/gossan; (2) a horizon that has been subjected to extreme acidification (acidified); (3) a supergene copper-enriched horizon; and (4) underlying primary massive sulphide mineralization.

The Bisha Main Zone is a 1.2 km long narrow massive sulphide lens that is oriented north-south. The true thickness of the lenses is variable from 0 to 70 m. The deposit is deformed and exhibits thickening at the fold hinge and limb attenuation, which distorts original dimensions. Drill intersections encountered mineralization to a depth of 380 m but portions of the deposit only extend to depths of 70 m. The mineralization remains open down dip in several portions of the deposit. Extensions at depth would add primary sulphide mineralization. Metal zoning within the massive sulphide appears to indicate an upward transition from Cu-rich to Zn-rich to barren pyrite.

Nevsun has completed a total of 50,715.76 m of drilling in 352 holes. Of this, 48,309.66 m was core drilling in 310 holes; 1,808.4 m was completed in 33 RC holes; and 591.70 m was in 9 combination holes, which had RC at the top of each hole and core drilling in the bottom.

Nevsun and contractors have completed work according to standard industry practices. A QAQC program was in place for the drilling program to monitor the accuracy and precision of the assays and ensure that sampling preparation and analytical protocols were being monitored.

AMEC also reviewed the QAQC program and verified the database used for geologic modelling and resource estimation. Results for the additional independent and check sampling collected by AMEC were acceptable. AMEC considers the practices used to collect the data to be in accordance with standard industry practices.

The deposit mineralogy is polymetallic and will require a significant testwork program to identify an optimum process flowsheet and to confirm metal products, grades and recoveries.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 20-1
	October 2004

Nevsun and AMEC completed geological modelling. A solid (wireframe) was developed for the geological model that outlined each principal mineralized or geological domain and became the basis for coding the block model.

Ken Brisebois, P. Eng., AMEC, prepared the mineral resource estimate with Vulcan 3D Software using industry standard methodologies conforming to the requirements set out in National Instrument 43-101. A total of 22.75 Mt was reported as Indicated mineral resources, and an additional 5.85 Mt was also reported as Inferred mineral resources (see Table 20-1) with varying cut-off depending on the primary metal for that domain.

Additional drilling will be required to advance the resource estimate to a feasibility study level. In Section 20.1 AMEC provides recommendations to advance the project towards engineering studies.

Other targets on the Bisha Property include the Northwest Zone, Harena, and the NW Barite Showing areas.

The Northwest Zone is interpreted to be a structural repetition of the stratigraphy hosting the Bisha Main Zone mineralization. Mapping, geochemistry, and geophysics all support this as a good exploration target with clear anomalies. Drilling intersected mineralization in 8 of the 14 holes and one of the longest intervals is from hole B-066 with a 47.5 m core length interval averaging 1.32 g/t Au, 14.96 g/t Ag, 1.52% Cu, 0.01% Pb, and 0.04% Zn (pers. comm. Nielsen, 2004). The true width of the interval is not yet known because the interpretations are preliminary. This zone warrants additional drilling.

The Harena Area is 9 km southwest of the Bisha Main Zone and has a gossan with associated geochemical and geophysical anomalies. The target warrants additional exploration including drilling.

The NW Barite Showing has a geochemical and geophysical anomaly. The target warrants additional exploration including drilling.

Nevsun has temporarily suspended fieldwork in response to a stop work order from the Government of Eritrea issued to all exploration companies. The circumstances of the order are not public.

20.2

Recommendations

During the site visit and review AMEC has made several recommendations for improvements to the work being conducted on the Bisha Property. Nevsun has already addressed many of the recommendations (i.e. corrections to database, change in magnetic declination, etc.). Other items that should be addressed are as follows:

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 20-2
	October 2004

• Implement a double data entry system or a data entry system with some form of validation of codes. Enter of data could be direct to MS Access or a relational database with filters, limits, and data integrity checks.
  
• Logging code system should be simplified, standardized and a set of equivalent codes should be prepared to equate the logging to surface mapping.
  
• Include a commercial blank sample within the QA/QC program. If the coarse blank material is also continued then care should be exercised during collection of the material.
  
• Assay a series of pulp samples (check assays) at an external laboratory (approximately 5% is a normal recommended number of check assays). Also, continue to submit 1 in 20 samples to a second independent laboratory other than ALS Chemex.
  
• Check the relative coverage of the bulk density measurements. If sufficient samples can be collected from earlier drilling or current drilling then assess the possibility of developing a density model for resource estimation.
  

AMEC considers that further drilling and engineering studies are warranted on the Bisha Property to advance the Bisha Deposit towards feasibility studies. The recommended activities include:

• Geotechnical assessment of the potential open pit parameters. The available geotechnical data, including oriented core should be reviewed and modelled. Geotechnical data collection involving drilling into the walls of the potential open pit will be required. A geotechnical engineer should provide the coordinates for geotechnical drilling and a series of procedures for logging and sampling of geotechnical information.
  

• Metallurgical samples should be collected and tested. The suite of samples must be representative of the mineralized domains and consider the specialized testwork required each area of the deposit. Sample collection will require large diameter core (PQ - 85 mm diameter) to provide an adequate volume and weight for metallurgical testwork. A metallurgical testwork program should be designed to assess process options and provide sufficient information for preliminary and/or feasibility studies. A reputable metallurgical testing facility needs to be retained to complete and report on this work.
  

• Baseline environmental studies, including social and archaeological studies (underway), in preparation for an Environmental Impact Assessment (EIA).
  

• Socio-economic assessment (underway).
  

• Hydrological studies (underway).
  

• Tailings containment system design (underway).
  

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 20-3
	October 2004

• Waste and tailings acid generation assessment (underway).
  

• Complete drilling to infill the near surface resources of the Bisha Main Deposit. Nevsun estimates that this will involve approximately 1,500 m of core drilling. To improve drill recoveries in the gossan area the holes should start with PQ-sized core and subsequently reduce to HQ and NQ as the competency of the rock increases with depth and the effects of surface weathering and oxidation diminish.
  

• Complete closer-spaced drilling on sections 12.5 m apart to provide improved confidence and advance a significant portion of the resources to the Measured category.
  

• Review the model and prepare an updated resource estimate including the mineralization in the stockwork and disseminated zones.
  

• Improve the roads and basic infrastructure in the immediate Bisha area.
  

All drilling and exploration work should include the QAQC programs and practices that are already in place. Nevsun and AMEC also consider that the following exploration activities should be completed:

• Deeper drilling on the Bisha Main Zone to determine the overall extent of the VMS deposit, i.e.:
  

-

down-plunge on the south end of the Main Zone.

-

along trend of mineralized horizon at north end of the Main Zone.

-

down-dip of eastern limb.

• Drill the Bisha Northwest VMS Zone.
  

• Drill the prime exploration targets such as the coincident gravity, HLEM and soil anomalies defined on the Harena Area (SW grid).
  

• Model the gravity targets south of line 1,715,000N using available specific gravity data from various core intervals to model the depth of any potential massive sulphide mineralization that may be present in this area.
  

• Conduct exploration and possible drilling in the NW Barite Showing area.
  

20.2.1

Phase I Work Program

Nevsun has provided a budget to address Phase I the activities described above which include:

• Geotechnical testwork.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 20-4
	October 2004

• Metallurgical testwork.
  

• Environmental assessment (ongoing studies).
  

• Archaeological assessment (ongoing studies).
  

• Tailings containment system design (ongoing studies).
  

• Waste and tailings acid generation assessment (ongoing studies).
  

• Socio-economic assessment (ongoing studies).
  

• Infill drilling near surface around the gossan.
  

• Drilling around the Bisha Main Zone.
  

• Update resource estimates.
  

• Improve roads.
  

• Exploration and drilling additional of targets listed above.
  

The total budget for Phase I is US$2.76 M. Exploration activities are approximately US$2.01 M (72%), engineering studies and testwork comprises approximately US$0.435 (16%), and camp operation and support is approximately US$0.313 M (11%) of the budget.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 20-5
	October 2004

Table 20-1
Phase I Work Plan and Budget

Bisha Main	Month (US$)						Total
Northwest and Exploration	1	2	3	4	5	6	US$
Diamond Drilling				$630,000	$810,000	$120,000	$1,560,000
Geophysics				$40,000			$40,000
Consultants	$20,000	$25,000	$25,000	$30,000		$30,000	$150,000
Eritrean wages	$1,500	$1,500	$5,000	$10,000	$10,000	$5,000	$33,000
Local Wages	$1,000	$1,000	$2,000	$5,000	$5,000	$5,000	$19,000
Food	$2,000	$2,000	$2,000	$25,000	$25,000	$25,000	$81,000
Fuel			$27,000	$30,000	$30,000	$15,000	$102,000
Construction	$2,000	$2,000	$2,000	$2,000	$2,000		$10,000
Supplies			$5,000	$5,000	$5,000	$5,000	$20,000
Vehicles				$1,000	$1,000	$1,000	$3,000
Assaying				$50,000	$50,000	$50,000	$150,000
Air fares				$35,000			$35,000
Communications			$30,000	$5,000	$5,000	$5,000	$45,000
Metallurgical				$100,000	$50,000	$50,000	$200,000
Reports	$20,000	$20,000				$25,000	$65,000
Satellite Imagery			$10,000				$10,000
Hydrological study				$40,000			$40,000
Geotechnical study					$25,000		$25,000
Environmental studies			$10,000	$60,000			$70,000
Resource estimate				$25,000	$25,000	$50,000	$100,000
Total	$46,500	$51,500	$118,000	$1,093,000	$1,063,000	$386,000	$2,758,000

20.2.2

Phase II Work Program

Furthermore AMEC considers that the Bisha Project should be advanced towards completion of a feasibility study.

AMEC recommends that the initial studies competed in Phase I will be used to develop a Scoping Study to identify the overall conceptual project scope and address: geology, mining, process, ancillary facilities, infrastructure, environmental, opportunities, risks, capital costs, operating costs and financial analysis. Costs to complete a Scoping Study could be US$0.100 to 0.200 M including the additional investigations required for preliminary mine plan, flowsheet, cost estimates etc. and the number of options considered.

The Scoping Study may show that additional studies are required or it may conclude that the next step is to commence a detailed Feasibility Study.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 20-6
	October 2004

21.0

REFERENCES

Abdelsalam M.G., Stern, R.J.(2001): Geology of the Neoproterozoic Basement around the Red Sea, in Rupturing of Continental Lithosphere in the Red Sea/Gulf of Suez, MARGINS Education and Planning Workshop, Sharm-El-Sheikh, Sinai, Egypt.

Barrie, C. Tucker (2004): Report on Geology and Geochemistry for the Bisha VMS Deposit and Property, Western Eritrea – An Internal Report for Nevsun Resources, August 2004.

Barrie, T.C., Hannington, M.D. (1999): Volcanic-Associated Massive Sulphide Deposits: Processes and Examples in Modern and Ancient Settings, Reviews in Economic Geology, Volume 8, Soc. Of Econ. Geol. & Geol Assoc of Canada. 408pg.

Berhe, S.M. (1990): Ophiolites in Northeast and East Africa: Implications for Proterozoic crustal growth, J. Geol Soc. London, 147, p. 41-57.

Chewaka, S. and DeWit, J.(1981): Plate Tectonics and Metallogenesis: Some Guidelines to Ethiopian Mineral Deposits, Bulletin #2, Ethiopian Institute of Geological Surveys, Min. of Mines, energy and Water Resources, Provisional Military Government of Socialist Ethiopia, 129pg.

Childe, F. (2003): Geological Model for Volcanogenic Massive Sulphide Mineralization on the Bisha Property Eritrea, contractors private report for Nevsun Resources Ltd by imp Interactive Mapping Solutions Inc. April 2003.

Chisholm, R., Delisle, P.C., Nielsen, F.W., Daoud, D., Ansell, S., Davis, G. (2003): Exploration and Drilling Program on the Bisha Property for Nevsun Resources (Eritrea) Ltd., Bisha Exploration Permit 2003, Work Program, August 2003.

Chisholm, R., (1999), Bisha Prospecting License, Gash-Barka Administration Region, Private company report for Nevsun Resources (Eritrea) Ltd, June 1999.

Drury, S.A. and Berhe, S.M. (1993): Accretion Tectonics in Northern Eritrea, Revealed by Remotely Sensed Imagery; Geol. Mag. 130 (2), pp.177-190, Cambridge University Press.

Drury, S.A. and Charlton, R. (1990?): The State of Eritrea (A Satellite Image Map) 1:1,000,000 scale LANDSAT mosaic of 14 images taken in 1985 and 1989, Western Lowlands dates from September; Open University, UK and Eritrean Latteragency Consortium.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 21-1
	October 2004

Franklin, J. M., Hannington, M. D., Jonasson, I. R. and Barrie, C. T. (1998): Arc-Related Volcanogenic Massive Sulphide Deposits; in Metallogeny of Volcanic Arcs, B. C. Geological Survey, Short Course Notes, Open File 1998-8, Section N.

Franklin, J. M., Lydon, J. W. and Sangster, D. F. (1981): Volcanic-associated Massive Sulphide Deposits. Economic Geology 75^th Anniversary Volume pp 485-627.

Fugro Airborne Surveys (2003) Logistics and Processing Report Airborne Magnetic and GEOTEM Survey, Bisha Area, Gash Barka District, Eritrea Job no.03427, Contractor’s private report.

Greig, C. (2004): Geology Report on Bisha Property, 2004.

Hðy, Trygve (1995): Noranda/Kuroko Massive Sulphide Cu-Pb-Zn, in British Columbia Geological Survey, 1995.

Leitch, C.H.B., (2003a): Petrographic Report on Four Polished Thin Sections for Nevsun Resources, Feb 09, 2003, 8 pages.

Leitch, C.H.B., (2003b): Petrographic Report on 11 Polished Thin Sections for Nevsun Resources, March 2003, 14 pages.

Leitch, C.H.B., (2003c): Petrographic Report on 13 Polished Thin Sections for Nevsun Resources, April 09, 2003, 26 pages.

Leroy, J-C (1989): Geology of Sahel Region, Eritrea; 1:450,000 (?) scale map drawn from the map by the Geological Dept. Of Eritrea, Peoples Liberation Front, Eritrea.

Mercier, M. (2003): Geochemical Surveys, A Final Report Bisha and Okreb Prospecting License Areas, Gash Barka Administrative Region, Eritrea, Contractor’s private report, June 2003.

Nevsun (2004): Exploration Program on the Bisha Property, Gash-Barka District, Eritrea 2004, Private company report for Nevsun Resources (Eritrea) Ltd., September, 2004.

Nevsun (2003): Exploration Program on the Bisha Property, Gash-Barka District, Eritrea 2002 for Nevsun Resources (Eritrea) Ltd., May 2003.

Nielsen, F.W., Chisholm, R.E., Woldu, A. (2003): Exploration Program on the Bisha Property, Gash-Barka District, Eritrea, 2002, Private company report for Nevsun Resources (Eritrea) Ltd, May 2003.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 21-2
	October 2004

McNabb, K. (2003): Logistical Summary, Bisha Area Gravity Survey, Eritrea, contractor’s private report for Nevsun Resources (Eritrea) Ltd by MWH Geo-Surveys, Inc., April 2003, 36 pgs.

Singer, Donald A. and Mosier, Dan L. (1986): Grade and Tonnage Model of the Kuroko Massive Sulfide, 1986.

United States Department of the Interior (1990): Mineral Industries of Africa, Minerals Yearbook, Volume III, 1990 International Review, Bureau of Mines.

Yager, Thomas R. (2000): The Mineral Industries of Djibouti, Eritrea, Ethiopia, and Somalia, 2000.

Sub-Sahara Resources N.L. (2004): www.subsahara.com.au CIA, 2004: www.cia.gov/cia/publications/factbook/geos/er.html ALS Chemex, 2004: www.alschemex.com Ma’ aden web site : www.maaden.com.sa

Alemu, T. 2002 : www.utdallas.edu/~milles/SedGeoChem/snowball.html

Alemu, T. (2004): Tectonic Evolution of the Pan-African Tulu Dimtu Belt : Implications for the Precambrian Geology of Western Ethiopia, 2004.

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Page 21-3
	October 2004

	Technical Report on the Bisha Property and Resource Estimate of the Bisha Deposit
	Project No. 145103	Appendices
	October 2004