EX-99.1 2 exhibit99-1.htm TECHNICAL REPORT DATED SEPTEMBER 30, 2011 Exhibit 99.1
Exhibit 99.1


NI 43-101 Technical Report
Dolores Gold-Silver Project
Chihuahua, Mexico

 



Prepared for:
Minefinders Corporation Ltd.

 



30 September 2011
117119



Prepared by:
Chlumsky, Armbrust & Meyer, LLC



Robert L. Sandefur P.E.

 




TABLE OF CONTENTS

        Page No.
Section        
1.0 SUMMARY   1
  1.1 Introduction 1
  1.2 Property   1
  1.3 History   2
  1.4 Geology and Mineralization 2
  1.5 Exploration, Sampling, and Assaying 3
  1.6 Mineral Resource Estimation 4
  1.7 Mineral Reserve Estimation 9
  1.8 Mine Operations 10
  1.9 Conclusions 10
  1.10 Recommendations 10
2.0 INTRODUCTION 12
3.0 RELIANCE ON OTHER EXPERTS 13
4.0 PROPERTY DESCRIPTION AND LOCATION 14
5.0 ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY 17
6.0 HISTORY   18
  6.1 Production History 20
7.0 GEOLOGICAL SETTING AND MINERALIZATION 22
  7.1 Regional Geology 22
  7.2 Local Geology 24
  7.3 Mineralization 27
8.0 DEPOSIT TYPES 30
9.0 EXPLORATION 31
10.0 DRILLING   32
11.0 SAMPLE PREPARATION, ANALYSES AND SECURITY 33
  11.1 QA/QC on Drilling Since Last 43-101 33
12.0 DATA VERIFICATION 34
13.0 MINERAL PROCESSING AND METALLURGICAL TESTING 37
14.0 MINERAL RESOURCE ESTIMATES 38
  14.1 Basis of Estimation 38
  14.2 Mineral Resources 39
    14.2.1 Geologic Interpretation 39
    14.2.2 Sample Databases 43
    14.2.3 Databases Provided to CAM 43
    14.2.4 Grade Estimation 44
    14.2.5 Block Model 56
    14.2.6 Tonnage Estimation 58
    14.2.7 Reconciliation 60
    14.2.8 Changes in the Resource Estimation Methodology and Impact of Changes 65
    14.2.9 Resource Tabulation 65
15.0 MINERAL RESERVE ESTIMATES 68
  15.1 Pit Optimization, Pit Design, and Economic Parameters 68
    15.1.1 Pit Optimization Parameters 68
    15.1.2 Pit Shell Optimization 70
    15.1.3 Ultimate Pit with Haul Roads 70
    15.1.4 Production Phases and Dump Location 71

 


i  

 




TABLE OF CONTENTS

      Page No.
Section      
    15.1.5 Internal Cutoff Calculation 71
  15.2 Reserve Summary 71
  15.3 CAM Review of Reserves 72
16.0 MINING METHODS 74
  16.1 Mining Operations 74
  16.2 Mine Equipment Fleet 74
  16.3 Drill & Blast 75
  16.4 Ore Control 75
  16.5 Mine Personnel 75
17.0 RECOVERY METHODS 76
  17.1 Process Flow Sheet 76
  17.2 Actual Recovery Results 77
18.0 PROJECT INFRASTRUCTURE 79
19.0 MARKET STUDIES AND CONTRACTS 80
20.0 ENVIRONMENT STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT 81
  20.1 Permitting 81
  20.2 Reclamation 82
21.0 CAPITAL AND OPERATING COSTS 83
22.0 ECONOMIC ANALYSIS 84
  22.1 Exploration Potential 84
23.0 ADJACENT PROPERTIES 85
24.0 OTHER RELEVANT DATA AND INFORMATION 86
25.0 INTERPRETATION AND CONCLUSIONS 87
26.0 RECOMMENDATIONS 88
27.0 REFERENCES 89
28.0 DATE AND SIGNATURE PAGE 91

 

ii  

 




TABLE OF CONTENTS

    Page No.
Tables    
1-1 Mineral Resources – Measured 7
1-2 Mineral Resources – Indicated 7
1-3 Mineral Resources – Measured + Indicated 8
1-4 Mineral Resources – Inferred 8
1-5 Dolores Mine Reserves as of 31-Dec-2010 9
2-1 Report Contributors and Responsibilities 12
4-1 Mining Concessions 15
6-1 Previous Mineral Resource Estimates for the Dolores Project 19
6-2 Dolores Mine Production 20
14-1 Identification of Domains 40
14-2 Minefinders Dolores 2011 Update Drilling Statistics from Assay MEDS 207 Database 43
14-3 Minefinders Dolores 2011 Update MF11 MDB Drilling Statistics from Assay Database 43
14-4 Minefinders Dolores 2011 Update MF11 MDB Drilling Statistics from 2m composite MDB Database 44
14-5 Capping Values 45
14-6 Threshold Check - NNHH vs. NNSH (NN Threshold vs. No Threshold) 47
14-7 Range Restricted Caps 52
14-8 Minefinders Dolores 2011 PROBABILISTIC Model Geometric Parameters 56
14-9 Dolores 2011 Constrained ID2 Model Geometric Parameters 56
14-10 Dolores 2011 CAM CHECK MODEL Geometric Parameters 57
14-11 Ranges and Rotations for Kriging Interpolation, High-Grade Zone Gold 57
14-12 Percent of MineSight  60
14-13 Reconciliation (Provided by Minefinders) For December 1,2010 Through March 31,2011 (4 Months)  63
14-14 RC assays vs. 2m Exploration Composites Bias test  64
14-15 Approximate (~) Statistical Uncertainty in Resource Estimates Based on 2m Composites in 4 months of Pit Progress  64
14-16 Mineral Resources – Measured 66
14-17 Mineral Resources – Indicated 66
14-18 Mineral Resources – Measured + Indicated  67
14-19 Mineral Resources – Inferred 67
15-1 Pit Optimization Parameters 68
15-2 Projected Metal Prices as of February 28, 2011 (CAM) 69
15-3 Costs Used in Cutoff Grade Calculations 71
15-4 Dolores Mine Reserves as of 31-Dec-2010 71
15-5 Dolores Mine Reserves as of 31-Dec-2007 72
16-1 Mine Equipment Fleet 74
17-1 Predicted Metallurgical Recoveries 76
21-1 Oustanding Sustaining Capital Costs 83

 

     
iii  

 




TABLE OF CONTENTS

    Page No.
Figures    
4-1 Dolores Location Map 14
4-2 Dolores Mining Concessions 15
7-1 Regional Geologic Map 23
7-2 Property Geology 25
7-3 Stratigraphic Column 26
9-1 Schematic Cross Section 29
14-1 Cross Section 1700 Domain 101 at left, Domain 102 at right. 41
14-2 Cross Section 2100 Domain 103 at left, domain 102 at right. 41
14-3 Cross Section 2200 Left to right: Domain 103, 104, 105, 102. East dike is isolated Domain 102 material at right. 41
14-4 Cross Section 2275 Left to Right: Domains 103, 104, 105, 102. Notice transverse shear zones in domain 104. 41
14-5 Cross Section 2425 Left to Right: Domains 103, 104, 105, 102. Notice transverse shear zones visible in domain 104 42
14-6 Cross Section 2625 Left to right: Domains 103, 106, 105 (yellow), 102. 42
14-7 Cross Section 2775 Domains: Left -103, Center - 104. Right - 105. 42
14-8 Cross Section 2800 Domains: Left -103, Center - 104. Right - 105. 42
14-9 Cross Section 3075 Domains: Left -103, Center - 104. Right - 105. 42
14-10 NN Volume Check for 0.38 High-Grade Indicator Threshold 48
14-11 NN Volume Check for 0.40 High-Grade Indicator Threshold 49
14-12 Cross Section 1925 50
14-13 Cross Section 2675 50
14-14 Cross Section 2225 51
14-15 Domain 101 – Au Variogram 54
14-16 Domain 102 – Au Variogram 54
14-17 Domain 103 – Au Variogram 54
14-18 Domain 104 – Au Variogram 54
14-19 Domain 105 – Au Variogram 54
14-20 Domain 106 – Au Variogram 54
14-21 Domain 101 – Ag Variogram 55

14-22

Domain 102 – Ag Variogram 55
14-23 Domain 103 – Ag Variogram 55
14-24 Domain 104 – Ag Variogram 55
14-25 Domain 105 – Ag Variogram 55
14-26 Domain 106 – Ag Variogram 55
14-27 MineSight Probabilistic Model – Swath Plot 62
14-28 Constrained ID2 Model – Swath Plot 62
15-1 Graph of Whittle LG Pits 70
17-1 Cumulative gold recovery, phase 2 leach pad, relative to modeled gold recovery 77
17-2 Cumulative silver recovery, Phase 2 leach pad, relative to modeled silver recovery 78

 

iv  

 




1.0 SUMMARY

 

1.1 Introduction


This report was prepared at the request of Minefinders Corporation Ltd. (herein "Minefinders"), a public Canadian company, to update the mineral resources and mineral reserves at the Dolores gold-silver mine in Chihuahua, Mexico, in compliance with Canadian National Instrument 43-101. This is the first update of reserves and resources for Dolores since the Technical Report of Gustavson Associates in 2008, which disclosed mineral resources and reserves as of 31 December, 2007.

Proven and probable reserves have been estimated as of December 31, 2010. Minefinders personnel prepared these estimates, with independent review by Robert L Sandefur, P.E. of Chlumsky, Armbrust and Meyer, LLC. Mr. Sandefur is the Qualified Person as defined by NI 43-101 for the mineral resource and reserve statements. He visited the Dolores project on 23-24 September, 2010.

1.2 Property


The Dolores project is located in the Sierra Madre Occidental Range in the State of Chihuahua, northern Mexico at Latitude 29° 00’ North, Longitude 108°32’ West. It is in the Municipality of Madera, about 94 km by road (45 km by air) southwest of the town of Madera, and 250 km west of the city of Chihuahua. The area is rugged, with elevations ranging from 1,200 to 2,000 meters. Vegetation ranges from thorn scrub with cacti, to oak and pine forests at higher elevations. Some snowfall occurs in winter, but exploration and mining work may be carried out year-round.

Minefinders, through its wholly owned Mexican subsidiary Compañía Minera Dolores, S.A. de C. V. (CMD), directly owns 100% of the three exploitation concessions: Dolores, Silvia and Unificacion Real Cananea. Total area of the three concessions is 27,700 hectares. Minefinders provided a property title opinion dated 20 December 2010, prepared by Mexican law firm Garcia-Jimenez & Asociados of Mexico City, certifying that the mineral titles are in good standing. The three concessions contiguous and include the entire Mineral Resource defined in this report. A Net Smelter Return (NSR) of 2 percent on silver and 3.25 percent on gold production is payable to Royal Gold, Inc. only on that portion of the Dolores deposit contained within the Unificacion Real Cananea concession. All fees and taxes for property owned or under the control of Minefinders have been fully paid and are up to date.

Much of the surface rights on the Dolores project are owned by Ejido Huizopa. An "ejido" is a uniquely Mexican institution set up by the government during a period of land reform in the early 1900’s. Minefinders have surface-right agreements with Ejido Huizopa and with several individual members of the ejido, allowing Minefinders access and the right to carry out exploration and mining activities. The agreement grants Minefinders irrevocable use of land for the mine and roads for a term of 15 years with a

1  

 




right to extend for 15 more years. The project includes access to sufficient surface rights for mining operations, tailings storage areas, waste disposal areas, heap leach pad areas and processing plant sites.

The old village of Dolores occupied an area directly over the southern end of the mineral resource. An agreement between Minefinders and Ejido Huizopa provided for relocation of the village. Construction of the new town is complete, and relocation of all families in the project area was completed at the end of 2009.

1.3 History


Lode mining in the Dolores region started by 1898, continuing through 1931, with incomplete records from 1922 to 1931 indicating production during this period of over 116,000 ounces of gold and 6,000,000 ounces of silver. This production includes some from outside the Dolores resource study area.

The property essentially remained idle until 1993 when Minefinders acquired property in the district. Exploration during 1995 to mid-2007 included about 200,000 meters of drilling in 850 holes, both core and reverse circulation.

Pre-stripping activities began in February 2008. Initial production was in November 2008, and Minefinders started commercial production in May 2009.

1.4 Geology and Mineralization


Gold and silver mineralization at Dolores occurs as low- to medium-sulfidation, epithermal gold-silver bearing veins, silica stockworks, breccias and replacements. The system is mostly structure-controlled, within a NNW striking extensional fault system. Epithermal deposits with features discussed below are present through the Sierra Madre Occidental of Mexico, in numerous productive districts. The mineralization and host rocks are of Tertiary age.

Mineralization occurs as fine-grained quartz-adularia veins, breccia fill and stockworks in the competent andesite flows of the Lower Volcanic Series. Relatively deep mineralization tends to be preferentially in high-grade veins. At higher elevations these feeder veins grade into wider stockworks, veinlet and disseminations toward the less competent, more permeable, overlying latite flows and tuffs of the Lower Volcanic Series. Near the surface, mineralization shows a strong element of structural control, but mineralization widens out owing to development of breccia and fractures adjacent to the main mineralized conduits. Steam-heated clay-illite-hematite alteration with no significant gold or silver values also is present at the surface and is a characteristic feature of this type of epithermal mineralization.

2  

 




The mineralized area at Dolores is over 4,000 meters long and up to 1,000 meters wide at elevations from 1,100 meters to 1,700 meters above sea level. Drilling has shown mineralization has to extend for over 700 meters of vertical extent, and full vertical extent has not been completely defined.

Mineralization is generally associated with quartz and may be composed primarily of iron oxides, silver sulfosalts, electrum, and native gold where oxidized grading into pyrite, silver sulfides, native silver and visible gold with increased galena, and sphalerite at depth. Other associated gangue minerals include sericite, illite, rhodochrosite, fluorite, chlorite, epidote and various secondary minerals. The mineralization is accompanied by quartz-adularia alteration. Locally pervasive silica alteration is closely associated with structures and may contain up to four percent fine-grained pyrite. Quartz textures are fine-gained to drusy encrustations in breccias. Areas of intense propylitic alteration are high in gold and silver values in some areas. Grades in the feeders are in the ranges of 10 to 30 g/t gold and 300 to more than 1,000 g/t silver. Disseminated mineralization that lies adjacent to high grade feeder structures at upper levels has lower grades in the range of 0.3 to 2 g/t gold and up to 150 g/t silver.

Some vein and disseminated gold-silver mineralization extends to the NNW and SSE, beyond the limits of the Resource estimate area, and these areas contain significant exploration potential.

1.5 Exploration, Sampling, and Assaying


Since 1993 all exploration work, including mapping, sampling, drilling and geophysical studies have been under direct supervision of Minefinders personnel. No assays determined prior to 1993 on the property have been used in preparation of this report. Since November 1995, the property has been subject to extensive geological, geochemical and geophysical surveys and exploratory drilling to identify drilling targets to define a mineralized gold and silver system.

Drilling prior to 2008 was adequately described in previous Technical Reports (e.g. Gustavson, 2008), to which report the reader is referred. Drilling and logging procedures have not changed materially since 2007. Recent drilling at Dolores has been oriented toward expanding the resource below and peripheral to the open pit mine. Additional in-resource drilling since 2007 has totaled 125 holes, in addition to the 803 holes used in the 2008 report. Only 29 of the additional 125 holes, were used in resource estimation, the remainder being geotechnical, outside the resources area, or otherwise excluded. Assay results from some of the additional holes were not available at the time of preparation of the resource estimate.

No changes to the sample preparation, analysis procedures or security have occurred since the previous Technical Reports. The data prepared and presented by Minefinders were audited by at least three other independent engineering firms. Only 13.5% of the drilling footage is material and not previously audited.

3  

 




There were no major post-2007 changes in the QA/QC routines and no changes in the drilling or sampling routine for the new drillholes. However, there has been no formal internal report generated to date which covers the QA/QC on the new drilling. There have been checks made to duplicates, blanks, and standards during data review, but there has as yet been no formal report generated covering this period. CAM requested but did not receive the QA/ QC data on the new drilling. However, based on previous work by Minefinders and the fact that the new drilling is less than the 15% of the total material drilling used in the resource estimate, CAM believes it is acceptable to use this data without independent review.

The sample database used for the resource has been maintained by Minefinders Exploration group using industry-standard GEMS software. Assay data are acquired and loaded electronically, while geologic information is hand-keyed into spreadsheets prior to loading.

CAM uses automated data processing procedures as much as possible in constructing and auditing geologic databases to assure consistency and minimize errors and costs. These procedures depend heavily on consistent alphanumeric attribute codes and consistent and non-duplicated field labels and drillhole IDs. While many of the issues flagged by these automated procedures are obvious to a human, CAM requires a clean and consistent database before proceeding with geological modeling. On the basis of these statistical checks, and the checks done in the past, CAM believes that the Dolores exploration database has been prepared according to industry norms and is suitable for the development of geological and grade models. CAM’s standard check procedures did find both errors and possible errors. The results were provided to Minefinders, although none of these issues or possible errors would have an overall impact on the global resource.

1.6 Mineral Resource Estimation


Mineral Resources were been estimated as of December 31, 2010. Minefinders personnel, principally Thomas Matthews, prepared these estimates, with independent review by Robert L Sandefur, P.E. of CAM. Mr. Sandefur is the Qualified Person as defined by NI 43-101 for this mineral reserve statement.

The mineral Resources and Reserves in this estimate were calculated using the Canadian Institute of Mining, Metallurgy and Petroleum (CIM), CIM Standards on Mineral Resources and Reserves, Definitions and Guidelines prepared by the CIM Standing Committee on Reserve Definitions and adopted by CIM Council on Dec 11, 2005.

In the past, the Dolores model had been estimated within hand-drawn high-grade and low-grade envelopes. The current model dispenses with the time-consuming manual sectional envelope construction

4  

 




process and uses a probabilistic approach for differentiating high grade from low grade zones for estimation.

Since production was initiated in 2009, Minefinders has used the common mining practice of determining which portions of the deposit are ore versus waste, based on the assays from vertical blast holes. Given the sub-vertical nature of the ore structures, it is not surprising that methodology did not yield entirely satisfactory results. Hence, in October 2010 an RC drilling program for grade control was initiated and fully implemented on December 1, 2010. In addition to the change from vertical blast holes to inclined RC holes for ore control, the mining methods were changed to mine perpendicular to the sub-vertical structures and geologists were assigned to monitor actual mining operations. CAM is in complete agreement with these changes.

In 2010, Minefinders developed a new resource model using geostatistical techniques for the sub-vertical ore zones are defined by indicators based on grade and then kriging is applied inside the ore zones to estimate the grade of the blocks. This multi-parameter methodology is becoming more common for resource estimation, particularly if actual production data are available. However, having so many parameters may make it possible to tune the model to actual production very closely and this may generate a false sense of accuracy with the model. Also, there is risk associated with holes bottomed in ore and the fact that usually the lower portions of the orebody are not drilled as densely as the upper part.

The use of actual production data to calibrate the model is complicated by the fact that Minefinders has recently made significant changes in the mining method and in ore control drilling and sampling. These changes, with which CAM is in complete agreement, better allow for the fact that the ore occurs in high grade sub-vertical structures which trend to the northwest.

The probabilistic resource model represents a significant change in the methodology used for the resource estimate at Dolores. In the past, the sub-vertical structures which contain ore grade material were manually interpreted and reviewed. This prior methodology, while accepted engineering practice, is both time-consuming to develop initially and time-consuming to update. This type of model gives an undiluted resource with no ore loss or dilution. Therefore, the appropriate ore loss and dilution factors must be applied to this type of resource model prior to its conversion to a reserve.

Structural Domains

Solids models were constructed to represent each of 6 structural domains for the resource model. These domains are intended to represent areas of distinctive structural geometry, wherein the primary structural trends which control mineralization are considered to be consistent and separable.

5  




Only drillhole information was used for estimation of block grades. The drill database comprises information from approximately 403 RC and 597 core holes. CAM was provided with the MineSight directory for the project (which contains the exploration database) and a Microsoft access file of the exploration database.

Capping values were set for each of the six domains, after consideration of several capping methodologies. Cap values ranged from 15 to 22 grams/tonne Au and from 500 to 1,000 grams/tonne Ag.

Drill data were composited in 5-meter intervals, starting from the top of the hole. Composites comprised entirely of void space are not assigned a grade (null value). An additional range-restricted capping of composites was employed during interpolation.

All composites above an indicator value were assigned a value of 1 and all composites below the value are assigned a 0. These indicators are then kriged. The resulting value gives the probability that the block should fall within the described zone. For the low-grade zone, a probability threshold of 0.150 (AuEQ) gpt was used. For the high-grade zone, a probability threshold of1.0 gpt (AuEQ) was used, as it corresponds to a second natural break in the data.

Grades were estimated using MineSight software. Grade interpolation is done separately for gold and silver. Interpolation is done by ordinary kriging using different search ellipses and nugget values for gold and for silver, as well as for each structural domain, for a total of 24 interpolation passes. Interpolation uses a minimum of 3 and a maximum of 8 composites for estimation.

Bulk density for all blocks is defaulted at 2.55 t/m3, the weighted average of bulk-density measurements for andesite (2.57 t/m3) and latite (2.53 t/m3). There is substantial complexity in the geometry of the latite dike swarm at Dolores and it was not been considered practical to create solids for density tagging of various rock types.

Block grades were estimated for all blocks below original mine topography. Resource and reserves reporting reports only include those blocks below the January 1 2011 surveyed topography surface from the mine. Underground workings have been encountered both in drilling and in mining, but no allowance was made to remove tonnes and grade associated with the old underground workings, as the tonnages in the workings are very small compared to the scale of the resource model.

Resource classification is based primarily on distance from data. A simplified interpolation ellipse was used, consistent with the scale of distance search used in the 2008 resource model, to flag blocks with distance to the closest composite. Blocks within a 15 x 20 x 17.5 search from composites were classed as

6  




measured, while blocks within a 30 x 40 x 35 search of composites were classed as indicated. More distal estimated blocks were classed as inferred.

Resource reporting is done at a series of gold-equivalent (75:1 ratio) cutoffs which roughly correspond to a range of probable cut-off grades. In actual practice the mine uses a recoverable dollar value cutoff to account for differences in recoveries for gold and silver, which vary according to alteration type. The current operating cutoff at the mine is $7.50 per tonne, although the best available material is always sent through the crusher first, which raises the nominal cutoff to $10 most months, with the lower grade material being stockpiled. This corresponds roughly to 0.3 gpt AuEQ grade.

To validate the total resource, CAM calculated contained tonnes, grade and contained ounces using a gold to silver ratio of 75 to 1. CAM reported resources between original surface topography and a pit surface provided by Minefinders based on measured, indicated and inferred classes. All of the models, except one, check within the plus or minus 5% which CAM regards as good. The exception checked within 7.7% in gold ounces, but for various reasons CAM considered the reconciliation to be good.

Resources tabulations are summarized in Tables 1-1 through 1-4. Note that because of the estimation method, the resource includes diluted ore reserves.

Table 1-1
Mineral Resources – Measured
Cut-off Tonnes Gold Gold Silver Silver Cut-off
(gpt AuEq)(i) (million) Ounces Grade Ounces Grade (gpt
    (million) (gpt) (million) (gpt) AuEq)(i)
0.2 96,587,826 1,605,129 0.517 83,718,850 27.0 1.036
0.3 80,050,610 1,530,279 0.595 79,241,979 30.8 1.187
0.4 61,857,757 1,415,775 0.712 72,508,982 36.5 1.414
0.5 46,836,549 1,293,629 0.859 65,502,085 43.5 1.696
0.6 36,734,441 1,194,358 1.011 59,652,930 50.5 1.982
1.0 21,765,502 989,672 1.414 48,087,568 68.7 2.735
1.5 14,558,872 823,659 1.760 39,004,084 83.3 3.362
2.0 9,291,428 652,792 2.185 83,718,850 99.9 4.106
(i) Gold-equivalent (AuEq) mine cut-off grades are based on a silver to gold ratio of 75 to one.
(ii) AuEq grade is based on a silver to gold ratio of 52 to one.

 

Table 1-2
Mineral Resources – Indicated
Cut-off Tonnes Gold Gold Silver Silver Cut-off
(gpt AuEq)(i) (million) Ounces Grade Ounces Grade (gpt
    (million) (gpt) (million) (gpt) AuEq)(i)
0.2 90,584,324 1,203,357 0.413 65,206,084 22.4 0.844
0.3 71,856,401 1,116,743 0.483 60,272,682 26.1 0.985

 

7  




Table 1-2
Mineral Resources – Indicated
Cut-off Tonnes Gold Gold Silver Silver Cut-off
(gpt AuEq)(i) (million) Ounces Grade Ounces Grade (gpt
    (million) (gpt) (million) (gpt) AuEq)(i)
0.4 52,033,568 988,672 0.591 53,231,017 31.8 1.203
0.5 37,090,620 865,371 0.726 46,434,526 38.9 1.474
0.6 28,057,187 775,032 0.859 41,331,407 45.8 1.740
1.0 14,534,863 589,634 1.262 30,790,123 65.9 2.529
1.5 9,219,244 463,983 1.565 24,506,244 82.7 3.155
2.0 5,671,306 351,683 1.929 18,134,789 99.5 3.842
(i) Gold-equivalent (AuEq) mine cut-off grades are based on a silver to gold ratio of 75 to one.
(ii) AuEq grade is based on a silver to gold ratio of 52 to one.

 

Table 1-3
Mineral Resources – Measured + Indicated
 Cut-off Tonnes  Gold Gold Silver Silver Cut-off
(gpt AuEq)(i) (million) Ounces Grade Ounces Grade (gpt
    (million) (gpt) (million) (gpt) AuEq)(i)
0.2 187,168,418 2,808,351 0.467 148,932,250 24.8 0.934
0.3 151,904,139 2,646,975 0.542 139,478,988 28.6 1.092
0.4 113,889,315 2,404,534 0.657 125,737,334 34.3 1.317
0.5 83,925,445 2,159,102 0.800 111,948,445 41.5 1.598
0.6 64,789,904 1,969,472 0.946 100,983,618 48.5 1.879
1.0 36,299,504 1,579,224 1.353 78,879,494 67.6 2.653
1.5 23,777,543 1,287,636 1.684 63,510,340 83.1 3.282
2.0 14,962,161 1,004,446 2.088 47,963,848 99.7 4.005
(i) Gold-equivalent (AuEq) mine cut-off grades are based on a silver to gold ratio of 75 to one.
(ii) AuEq grade is based on a silver to gold ratio of 52 to one.

 

Table 1-4
Mineral Resources – Inferred
Cut-off Tonnes Gold Gold Silver Silver Cut-off
(gpt AuEq)(i) (million) Ounces Grade Ounces Grade (gpt
    (million) (gpt) (million) (gpt) AuEq)(i)
0.2 40,786,301 388,139 0.296 19,459,389 14.8 0.581
0.3 27,632,397 326,214 0.367 16,133,009 18.2 0.717
0.4 17,359,931 256,959 0.460 12,764,245 22.9 0.900
0.5 11,089,833 202,514 0.568 10,118,552 28.4 1.114
0.6 7,451,842 164,086 0.685 8,219,877 34.3 1.345
1.0 2,755,033 96,750 1.092 4,788,327 54.1 2.132
1.5 1,323,119 62,140 1.461 3,223,125 75.8 2.919
2.0 668,668 40,646 1.891 2,087,209 97.1 3.758
(i) Gold-equivalent (AuEq) mine cut-off grades are based on a silver to gold ratio of 75 to one.
(ii) AuEq grade is based on a silver to gold ratio of 52 to one.

 

8  




Mineral resources which are not mineral reserves do not have demonstrated economic viability. The estimate of mineral resources may be materially affected by environmental, permitting, legal, title, taxation, socio-political, marketing, or other relevant issues.

The quantity and grade of reported inferred resources in this estimation are conceptual in nature.

1.7 Mineral Reserve Estimation

Mineral Reserves were estimated as of December 31, 2010. Note that because of the estimation method, the resource includes Diluted ore reserves are included within the resources reported above.

A standard methodology for pit limit analysis was followed by Minefinders using the Whittle software package which applies an implementation of the Lerchs-Grossmann (LG) algorithm. The parameters for pit optimization include a gold price of $1,200/oz and a silver price of $23/oz. Initial block dimensions were 3.0 m X 5.0 m X 7.5 m vertically.

The Whittle LG pit shell selected for the reserves statement corresponding to a revenue factor of 0.720, or $864 gold and $16.56 Silver. An ultimate pit plan including haul roads and safety berms was created using MineSight software by the Dolores planning engineering department. This ‘ultimate pit’ was used for reporting of reserves.

For the purpose of reporting mine reserves, the value of gold and silver in each block, net of recovery are stored (recoveries vary according to alteration for each block). An internal value cutoff grade (excluding mining cost) is used for separation of ore from waste based on the costs, which total $ 7.50/tonne.

CAM has reviewed the pit optimization parameters, pit design, cost parameters and cutoff calculations and finds them to be reasonable and acceptable. CAM reviewed the ultimate pit design and finds that it conforms acceptably to the Whittle LG pit shell (pit 29) upon which it was based. CAM verified the reserve quantities stated between the 31-Dec-2010 topographic surface and the ultimate pit design.

Minable reserves for the Dolores Mine as of 31-Dec-2010 are summarized in Table 1-5.

Table 1-5
Dolores Mine Reserves as of 31-Dec-2010
        Au Oz Ag Oz
Category Tonnes Au (g/t) Ag (g/t) (million) (million)
Probable 62,430,655 0.62 34.2 1.254 68.552
Proven 45,212,669 0.53 31.6 0.770 45.972
           
Total 107,643,324 0.58 33.1 2.024 114.524

 

9  




1.8 Mine Operations

The Dolores mine entered commercial production in April 2009. The operation moves approximately 90,000 tonnes of material per day and processes 17,500 tonnes of ore per day.

Mining is carried out by standard open pit methods, utilizing PC3000 Shovels, WD900 loaders, and 100-ton haul trucks. Ore control is drilling is carried out using angled RC drilling, with vertical blasthole drilling for blasting.

Ore is processed through a 3-stage crushing circuit producing material of 100% passing 3/8” crushed size. Ore is conveyed to the leach pads and stacked using grasshoppers and radial stacker. Gold and silver are extracted using sodium cyanide solution, and recovered in a Merrill-Crowe plant.

1.9 Conclusions

Based on a review of the database, resource estimation methodology and reconciliation provided by Minefinders and additional independent calculations by CAM, CAM believes that the resource model for Dolores has been prepared according to accepted engineering practice, and is suitable for public disclosure and for use in reserve calculations and financial planning.

There is only a limited amount of production data under the new mining method and the resource estimate may need to be revised as more production experience with the new mining method is gained.

1.10 Recommendations

 

1.

Correct errors identified by CAM in the Exploration and RC databases. Although it appears that these errors do not affect the global resource estimate they may be significant locally and should be corrected.

2.

The current procedure of reviewing QA/QC data for RC duplicates as a simple percentage difference is not best engineering practices as it generally tends to flag low grade samples and ignore errors at the higher grade. A simple scatter plot of original versus replicate value on both in untransformed and log-log basis should be adequate for this check.

3.

A formal report on QA/QC on the drilling since the last (2008) report needs to be prepared.

4.

The CAM sector search check model gave fewer ounces than either of the two Minefinders models which did not use a sector search. Using a sector search at least as a sensitivity run is a good check to avoid central higher grade composites possibly over driving the pit.

 

10  




5.

There are some cases (e.g. the x-coordinate for composites) where the maximum and minimum specified for the fields in MineSight are less or greater than the actual maximum and minimum of the data. This appears to have had no affect on the resource estimate but should be corrected.

6.

The multi-run feature in MineSight is a powerful and useful tool for developing complicated models. However, some multiple runs overwrite the report files from prior runs, rendering it impossible to properly audit the preparation the model. For this reason CAM recommends that each multi-run report file have a unique name.

7.

The probabilistic model lacks the tight constraints of the interpreted structures used in the constrained inverse distance squared model and has significantly wider across strike search than the CAM checked models. This means that there may be risk of a few high grade composites at the bottom of the hole local re-locally resulting in the prediction of too much metal. It is a relatively simple matter to calculate the amount of metal associated with each hole and composite particularly after recommendation six is implemented.

8.

As of the data close off date there was insufficient data to completely validate the model against the new model. After one year of production with the new ore control method, the reconciliation should be reviewed and the exploration model revised as necessary.

 

11  




2.0 INTRODUCTION

This report was prepared by Chlumsky, Armbrust and Meyer, LLC (herein “CAM”) at the request of Minefinders Corporation Ltd. (herein "Minefinders"), a public Canadian company, to update the mineral resources and mineral reserves at the Dolores gold-silver mine in Chihuahua, Mexico, in compliance with Canadian National Instrument 43-101. This is the first update of reserves and resources for Dolores since the Technical Report of Gustavson Associates in 2008, which showed mineral resources and reserves as of 31 December, 2007.

Proven and probable reserves have been estimated as of December 31, 2010. Minefinders personnel prepared these estimates, with independent review by Robert L Sandefur, P.E. of CAM. Mr. Sandefur is a Qualified Person as defined by NI 43-101 for the mineral resource and reserve statements.

The contributors to this report are enumerated in Table 2-1.

Table 2-1
Report Contributors and Responsibilities
Name Profession Affiliation Qualified Section Latest Visit to
      Person Responsibilities Property
Robert L. Sandefur P.E., Geostatistician Chlumsky, Armbrust and Meyer, LLC Yes overall, incl. 12, 14 23-24 Sep 2010
Richard L. Nielsen Ph.D., Geologist Chlumsky, Armbrust and Meyer, LLC Yes 7 and 8 5 – 7 July, 2007
Thomas Bagan V.P. Corp. Development Minefinders Corporation Ltd. Yes 17, Whittle shell April 2011
Laurence Morris* V.P. of Operations Minefinders Corporation Ltd. Yes parts of 16-21 August 2011
Thomas Matthews V.P. Information Services Minefinders Corporation Ltd. No Initial report assembly August 2011
* based on-site

The Dolores project was visited by Robert Sandefur on 23-24 September 2010, and previously by CAM geologist and QP Richard Nielsen in July, 2007. The geologic database has not changed significantly since 2007.

Some sections of this report are taken from previous Technical Report by CAM (2007) and by Gustavson (2008), with updates and edits as warranted.

All dollar amounts mentioned in this report are U.S. dollars.

12  




3.0 RELIANCE ON OTHER EXPERTS

This report includes information from four previous 43-101-compliant reports: Gustavson, 2008; CAM, 2007; Roscoe Postle, 2004; and Pincock Allen and Holt, 2002, as listed in the Reference section. Additional data and updates were made available by Minefinders, including a property title opinion dated December 2010, prepared by a Mexican legal firm, as discussed in Section 4.

The resource estimate was done by and under the direction of Thomas Matthews VP of Information Systems for Minefinders. Most of the resource section was also prepared by Thomas Matthews with some edits by CAM.

The reserve and mining section was prepared under the direction of Thomas Bagan, vice president of corporate development and Laurence Morris, VP of operations for Minefinders. The reserve and mining section was provided by Thomas Matthews.

Minefinders’ Canadian legal counsel, Stikeman Elliot, LLP, provided CAM with a letter dated 14 July 2011, stating that in their opinion, the information required under Section 22 is not required to be included in this Technical Report. Section 22 is therefore not completed in this report.

13  




4.0 PROPERTY DESCRIPTION AND LOCATION

The Dolores project is located in the Sierra Madre Occidental Range in the State of Chihuahua, northern Mexico at Latitude 29° 00’ North, Longitude 108°32’ West. It is in the Municipality of Madera, about 94 km by road (45 km by air) southwest of the town of Madera, and 250 km west of the city of Chihuahua (Figure 4-1).

Figure 4-1
Dolores Location Map

Mineral rights to the Dolores Project are controlled by three mining concessions shown in Table 4-1 and Figure 4-2, below. Minefinders, through its wholly owned Mexican subsidiary Compañía Minera Dolores, S.A. de C. V. (CMD), directly owns 100% of the Dolores, Silvia and Unificacion Real Cananea concessions. Total area of concessions under control of Minefinders remains 27,700 hectares. Minefinders provided a property title opinion dated 20 December 2010, prepared by Mexican law firm

14  




Garcia-Jimenez & Asociados of Mexico City, certifying that the mineral titles described below are in good standing.

Table 4-1
Mining Concessions
Dolores Mineral Concessions Title No. Area (ha) Expiry Date
Silvia 217587 2,866 August 20, 2052
Unificacion Real Cananea 227028 1,920 December 12, 2039
Dolores 221593 22,914 March 3, 2054
Total   27,700  

Figure 4-2
Dolores Mining Concessions

15  




A Net Smelter Return (NSR) of 2 percent on silver and 3.25 percent on gold production is payable to Royal Gold, Inc. only on that portion of the Dolores deposit contained within the Unificacion Real Cananea concession. There is no royalty on the other two concessions, which at this time do not contain reserves.

The three concessions shown on Figure 4-2 and listed in Table 4-1 are contiguous and cover the Mineral Resource defined in this study. All claims, license fees and taxes for property owned or under the control of Minefinders have been fully paid and are up to date. All concessions are of exploitation status. The claims are on file in the Municipality of Madera, State of Chihuahua, Mexico. During 2011, claim fees totaling approximately US$ 179,000 must be paid to maintain the claims in good standing.

Much of the surface rights on the Dolores project is owned by Ejido Huizopa. An "ejido" is a uniquely Mexican institution set up by the government during a period of land reform in the early 1900’s. It is a rural agricultural cooperative having well-defined property rights. Minefinders have surface-right agreements with Ejido Huizopa and with several individual members of the ejido, allowing Minefinders access and the right to carry out exploration and mining activities. The current agreement, signed in November 2006, grants Minefinders irrevocable use of land for the mine and roads for a term of 15 years with a right to extend for 15 more years. The agreement calls for cash payments, and other benefits to the ejido, including construction of schools, clinics, sports fields and other infrastructure, and preferential hiring for jobs and training to members of the ejido.

The project includes access to sufficient surface rights for mining operations, tailings storage areas, waste disposal areas, heap leach pad areas and processing plant sites.

The old village of Dolores occupied an area directly over the southern end of the mineral resource. An agreement between Minefinders and Ejido Huizopa provided for relocation of the village. Construction of the new town is complete, and relocation of all families in the project area was completed at the end of 2009.

The permits necessary to conduct work on property, and environmental/social issues, are discussed in Section 20 of this report.

16  

 




5.0

ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY


This section is taken from Gustavson (2008) with minor edits.

The property lies within the Sierra Madre Occidental mountains of northern Mexico, in the state of Chihuahua. Topography is rugged, with elevations ranging from 1,200 to 2,000 meters. Erosion has resulted in deep, V-shaped valleys and canyons. Vegetation is typical of the foothills of the Sierra Madre and ranges from thorn scrub with cacti, to oak and pine forests at higher elevations.

The climate is typical of the Sierra Madre with an average temperature of 18° C, and annual lows of -10° and highs of 45° C. Precipitation averages 250mm, most of it occurring from July to September as brief heavy rainstorms. Snowfall is common in December and January but does not remain on the ground for long. Exploration and mining work may be carried out year-round.

The main road access to the property is via 92 km of recently-upgraded dirt access road from Yepachi, Chihuahua, to the mine site. Access is also possible by light aircraft landing on a dirt strip located about 8 km from the mine.

The local economy is based on logging, ranching and subsistence farming. Unskilled workers may be found in nearby small villages. The company has a number of recruiting and training programs in place to develop the local workforce. However, the local population is small and the mine work force is supplemented from Chihuahua and Hermosillo, Sonora 350 km to the west. A strong culture of mining in both Sonora and Chihuahua states provides a pool of experienced workers.

Water is available from wells, from underground workings, pit dewatering activities, and from Rio Tutuaca. Additional water rights have been acquired for the mine, and a dam and reservoir were completed in August 2011 for storm-water control and primary water supply. Diesel generated power is generated on-site for production facilities.

17  

 




6.0 HISTORY

CAM and others have reported that mining in the Dolores region started with placer mining in the 1860’s followed by lode mining beginning by 1898. By 1915, a power line had been installed from the town of Madera, and a 25 tpd stamp mill began operation treating the Dolores ore. Pulverized ore was passed through a cyanide leach circuit with zinc precipitation for gold and silver recovery. The stamp mill operated until early 1929 when it was destroyed by fire.

Only sporadic high-grade production occurred from 1929 to 1931, with no records of any production since that time. Incomplete mining records from 1922 to 1931 indicate that total production during this time was about 372,000 tonnes of ore containing over 116,000 ounces of gold and 6,000,000 ounces of silver. This production came from several underground mine operations including some beyond the Dolores resource study area. In 1980, Consejo de Recursos Minerales, a Mexican government organization, sampled many of the accessible underground workings.

The property essentially remained idle until 1993 when Minefinders began acquiring a land position in the district. The Company initiated a full exploration project in November 1995, and drilling started in September 1996, with a total of about 200,000 meters of drilling to June 2007 with 850 holes of both core and reverse circulation drilling.

In July, 1996, Minefinders granted Echo Bay Mines (Echo Bay) an option to earn 60 percent interest in the Dolores Property by participating in a private placement of $5.7 million for the first phases of exploration, funding a bankable feasibility study, and paying Minefinders $20/oz on 60% of the proven and probable, gold and gold equivalent silver ounces. In October 1997, Minefinders bought-out the Echo Bay interest and owns the property outright subject to the underlying agreement. All data from Echo Bay’s complete drilling, sampling, environmental data collection, and metallurgical testing was transferred to Minefinders and in this report is not referenced separately as being undertaken by Echo Bay. Humboldt Mining Services prepared independent mineral resource estimates of the Dolores property in August 1997. In October 1998, Mineral Resources Development, Inc. (MRDI) of Canada completed a scoping study to evaluate development of the Dolores property. PAH conducted an audit of the Dolores resources in December 2002. Previous mineral resource estimates are listed in Table 6-1.

18  

 




Table 6-1
Previous Mineral Resource Estimates for the Dolores Project
Company Date of Resource Tonnes Au Ag AuEq
SEDAR filing Category (millions) (g/t) (g/t) (g/t)
Humboldt Mining
Services
Aug 1997
(pre-NI-43-101)
unclassified 27.8 1.3 47.3  
 
MRDI
0.5 g/t AuEq cutoff
Jun 1998
(pre-NI-43-101)
Indicated 30.2 0.852 41.3 1.40
Inferred 30.7 0.805 46.8 1.43
 
PAH Audit 43-101
0.3 g/t AuEq cutoff
Aug 2002 Measured 30.0 0.86 45.0 1.61
Indicated 41.5 0.75 47.7 1.55
Inferred 44.0 0.67 38.1 1.30
 
Roscoe Postle 43-101
0.3 g/t AuEq cutoff*
Nov 2004 Measured 53.4 0.89 43.7 1.48
Indicated 47.6 0.78 37.5 1.28
Inferred 28.1 0.77 28.0 1.14
 
CAM audit 43-101
0.3 g/t AuEq cutoff, 0.4
for inferred
Apr 2007 Measured 62.4 0.87 41.8 1.57
Indicated 60.9 0.76 36.2 1.37
Inferred 30.3 0.68 28.2 1.06
Gustavson Associates
43-101
same cutoff as CAM
audit, 2007.
March 2008 Measured same as CAM audit, 2007
Indicated
Inferred
 
*AuEq at 60:1 Ag/Au ratio

It should be noted that the Humboldt and MRDI resource estimates in 1997 and 1998 were undertaken prior to the establishment of NI-43-101 norms. Those resource estimates are historical in nature and as such are based on prior data and reports. The work necessary to verify the classification of the mineral resource estimates was not completed and the resource estimates therefore, cannot be treated as NI 43-101 defined resources verified by a qualified person. The historical estimates should not be relied upon and there can be no assurance that any of those resources, in whole or in part, will ever become economically viable. However, the subsequent (2002 and later) resource estimates were all 43-101 compliant.

In November 2002, M3 Engineering and Technology Corporation (“M3”) was contracted by Minefinders to initiate a series of scoping studies in support of a feasibility study. A complete bankable level feasibility study for the project was prepared in June 2005, based upon the results of earlier studies. KCA completed a separate bankable feasibility study in March of 2006 which forms the basis for the current project completion. Additional exploration drilling was completed in 2006 with the objective of obtaining data that would further enhance quality of the Mineral Resource estimate. CAM was engaged in

19  

 




May 2007 to prepare a new audit of the Minefinders revised Resource estimation that integrated the newly acquired data. Gustavson was retained in November of 2007 to prepare a reserve estimate and production schedule based on the audited Minefinders April 2007 mineral resource.

The only previous public mineral reserve estimate for Dolores is shown in Table 15-5.

Commencement of commercial production occurred in May, 2009. Details of the operation are discussed in Section 19 of this report.

6.1 Production History

The Dolores Mine production is summarized in Table 6-2.

Table 6-2
Dolores Mine Production
Item Units Year
2008 2009 2010
Mining
Ore Tonnes Mined Tonnes
(1000’s)
2,929 6,373 5,889
Waste Tonnes Mined (excl. pre-strip) Tonnes
(1000’s)
n.a. 19,779 12,346
Waste Tonnes Mined (pre-strip) Tonnes
(1000’s)
15,645 2,388 9,410
Total Tonnes Mined (incl. pre-strip) Tonnes
(1000’s)
18,574 28,540 27,645
Strip Ratio (incl. pre-strip) Waste:Ore 5.34 4.48 3.69
Strip Ratio (excl. pre-strip) Waste:Ore n.a. 2.79 2.10
 
Processing
Ore Tonnes Crushed Tonnes (1000’s) n.a. n.a. 5,549
Ore Tonnes Stacked Tonnes (1000’s) 1,124 5,546 5,554
Gold Grade Stacked grams/tonne 0.43 0.70 0.48
Gold Recovery Rate % n.a. n.a. 75.7
Gold Ounces Stacked (recoverable) Troy Ounces n.a. n.a. 64,642
Silver Grade Stacked grams/tonne 23.50 20.58 40.91
Silver Recovery Rate % n.a. n.a. 46.1
Silver Ounces Stacked (recoverable) Troy Ounces n.a. n.a. 3,371,348
AuEq Ounces Stacked (recoverable) Troy Ounces n.a. n.a. 121,255
 
Metal Production
Gold Ounces Produced Troy Ounces 2,440 77,264 56,110
Silver Ounces Produced Troy Ounces 42,800 1,318,245 1,218,663
AuEq Ounces Produced Troy Ounces n.a. n.a. 77,110

 

20  

 




Table 6-2 shows some decrease in gold and silver production in 2010 owing to loss of leachate on Pad 1 due to a tear in the liner. This incident is described more fully in Section 15.5 of this report.

21  

 




7.0 GEOLOGICAL SETTING AND MINERALIZATION

This section is taken largely from Overbay and others (2001), with minor edits by the present authors.

7.1 Regional Geology

The Sierra Madre Occidental (Figure 7-1) is a 1,200- by 300-km northwest-trending mountain system, and features a long northwest-trending volcanic plateau within a very broad anticlinal uplift. The region is dominated by large-volume rhyolitic ash flow tuffs emanating from calderas of Oligocene age (35 to 27 million years), the Upper Volcanic Series. These volcanic rocks are generally calc-alkalic rhyolitic ignimbrites with subordinate andesite, dacite, and basalt, with a cumulative thickness of up to 1,000 meters. The Upper Volcanic Series unconformably overlies rocks of the slightly older Lower Volcanic Series that are primarily of andesite composition with interlayered felsic ash flow tuff deposits that formed from 46 to 35 million years ago (Eocene time).

Deposition of the Lower Volcanic Series was accompanied by emplacement of hornblende-bearing quartz diorite and granodiorite batholiths and small intrusive bodies. The majority of the epithermal and porphyry related precious-metals deposits in the Sierra Madre Occidental are hosted in the Lower Volcanic Series.

Thin basalt to rhyodacite flows of late Miocene and younger age cap many of the plateaus and hills in the region. The Baucarit formation, a conglomeratic, basin fill sedimentary unit intercalated with several thin basalt flows, was deposited during Pliocene and Pleistocene time.

The oldest structural episode is related to the Laramide orogeny, which produced east-striking, steeply-dipping strike-slip faults. Later extensional forces resulted in development of N-S to N30W striking, sub-vertical, dip-slip normal faults of regional extent that produced a series of parallel to sub parallel west-dipping fault planes showing only limited horizontal displacement. Following these two events, NW-trending extensional forces resulted in development of N60E oriented normal faults.

Structures developed in the Dolores area are believed to have controlled emplacement of a series of NNW- trending andesite to latite intrusions. Zones of permeability associated with these faults and intrusive contacts formed conduits for the ascending mineralizing hydrothermal fluids.

22  

 




Figure 7-1
Regional Geologic Map

23  

 




7.2 Local Geology

The Dolores project is underlain by the two main Tertiary volcanic packages. The Lower Volcanic Series consists of gently tilted lavas, flow breccia, and tuffaceous rocks with a minimum thickness of 700 meters. It is conformably overlain by 100 to 200 meters of felsic latite volcaniclastic breccia. These units are overlain by the Upper Volcanic Series, which is a sub-horizontal volcaniclastic assemblage of mostly felsic ignimbrites and tuffs. The Upper and Lower Volcanic Series are separated by an unconformity (Figure 7-2) that is manifest by a distinctive polylithic, poorly-consolidated rubble zone of probable colluvial origin. This erosional surface formed after development of a NNW-trending anticlinal uplift. Subsequent erosion formed a window exposing mineralized rocks of the Lower Volcanic Series in the Dolores district.

The Upper Volcanic Series consists of felsic ignimbrites, volcanic breccias, tuffs and flows (Figures 7-2 and 7-3). Individual felsic units vary from stratified, poorly-lithified tuff to strongly-welded volcaniclastic rocks with interlayered tuffaceous conglomerate and breccia. All the units exposed are unsorted and range from non-stratified to stratified with monolithic to multilithic compositions. The Lower Volcanic Series (Figure 7-3) forms a varied package of volcanic rocks. Exposed rocks include light-colored coarse lithic lapilli tuff with fragments contained within a non-welded matrix, and latite flow rocks of porphyritic character with phenocrysts of K-feldspar and plagioclase.

24  




Figure 7-2
Property Geology

25  

 




Figure 7-3
Stratigraphic Column

26  

 




7.3 Mineralization

Gold and silver mineralization identified at surface at Dolores lies in an area over 4,000 meters long and up to 1,000 meters wide at elevations from 1,100 meters to 1,700 meters above sea level (Figure 7-2). Mineralization has been investigated to the depth of drilling at 1,000 meters elevation or over 700 meters of vertical extent (Figure 9-1). The extent of mineralization at depth has not been fully defined.

Mineralization in feeder structures at depth ranges up to five to ten meters wide in quartz stockworks and mineralized fractures. The bulk of the deposit is high in the system, where feeders widen into breccias and stockworks up to a few hundred meters wide that form halos around the main structures. Grades in the feeders are in the ranges of 10 to 30 g/t gold and 300 to more than 1,000 g/t silver. Disseminated mineralization that lies adjacent to high grade feeder structures at upper levels, has lower grades in the range of 0.3 to 2 g/t gold and up to 150 g/t silver. The width of coherent mineralization commonly is about 200 up to 300 meters wide, and lies in zones that trend NNW and dip steeply to the west. Some vein and disseminated gold-silver mineralization is known to extend in a NNW and SSE direction beyond the limits of the Resource estimate area, and this mineralization may be considered exploration potential and may eventually be included in the resource.

Breccia bodies range from true tectonic breccias, cemented by quartz, to hydrothermal breccias showing streaming textures. Breccias show evidence of episodic mineralization and re-cementation by quartz, adularia, and calcite.

Mineralization is generally associated with quartz and may be composed primarily of iron-oxides, silver sulfosalts, electrum, and native gold where oxidized grading into pyrite, silver sulfides, native silver and visible gold with increased galena, and sphalerite at depth. Other associated gangue minerals include sericite, illite, rhodochrosite, fluorite, chlorite, epidote and various secondary minerals.

Well-defined geochemical zonation patterns are characteristic with Sb, and As occurring high in the system and Pb, and Zn base metal mineralization increasing at depth. Combined percentages of Pb and Zn at elevations below the 1350 meter level locally exceed 3 to 10% and may be of future economic interest but are not included in this resource estimation.

Mineralization at Dolores is accompanied by quartz-adularia alteration. Locally pervasive silica alteration is closely associated with structures and may contain up to four percent fine-grained pyrite. Quartz textures are fine-gained to drusy encrustations in breccias. Areas of intense propylitic alteration are high in gold and silver values in some areas.

27  

 




Alteration zones include argillization along margins of the feeder structures, various intensities of propylitic (chlorite-epidote-calcite-pyrite) alteration of the volcanic wall rocks, as well as a regional low-grade propylitic overprint in the volcanic host rocks. The hydrothermal propylitic alteration related to mineralization is pyritic and relatively coarse grained. Patches of steam-heated alteration with kaolin-illite-hematite alteration and not carrying gold-silver values is present in the non-welded latite tuff breccias and flows of the Lower Volcanic Series. Oxidation of mineralization is highly variable throughout the deposit and has been noted to the full depth of drilling along structures. The deposit can be regarded as having variable amounts of oxide and mixed oxide sulfide grading into sulfide with depth.

Latite porphyry magmas are thought to be genetically related to emplacement of the epithermal veins. The regional structural system controls both emplacement of latite dikes and fluid flow related to mineralization. Strongest controls on mineralization are contact zones between intrusive latite and the andesite wallrocks where development of fractures and brecciation are focused. Mineralization is the result of ascending epithermal fluids that undergo cooling by fluid mixing, repeated boiling and throttling, and subsequent decompression and mineral deposition.

28  

 




Figure 9-1
Schematic Cross Section

29  

 




8.0 DEPOSIT TYPES

This section is taken largely from Overbay and others (2001), with minor edits.

Gold and silver mineralization at Dolores is present as low to medium sulfidation, epithermal gold-silver bearing veins, silica stockworks, breccias and replacements (Figure 9-1). The system is mostly structure-controlled, within a NNW striking extensional fault system. Epithermal deposits with features discussed below are present through the Sierra Madre Occidental as shown in Figure 4-1, and as characterized by Albinson and others (2001). Numerous productive districts are of this type.

Mineralization occurs as fine-grained quartz-adularia veins, breccia fill and stockworks in the competent andesite flows of the Lower Volcanic Series. Relatively deep mineralization tends to be preferentially in high-grade veins. At higher elevations these feeder veins grade into wider stockworks, veinlet and disseminations toward the less competent, more permeable, overlying latite flows and tuffs of the Lower Volcanic Series. Near the surface, mineralization shows a strong element of structural control, but mineralization widens out owing to development of breccia and fractures adjacent to the main mineralized conduits. Steam-heated clay-illite-hematite alteration with no significant gold or silver values also is present at the surface and is a characteristic feature of this type of epithermal mineralization.

30  

 




9.0 EXPLORATION

Since 1993 all exploration work, including mapping, sampling, drilling and geophysical studies have been under direct supervision of Minefinders personnel. No assays determined prior to 1993 on the property have been used in preparation of this report.

Minefinders initially carried out more than six square kilometers of detailed mapping and 12 square kilometers of reconnaissance mapping at Dolores. Geophysical surveys have included 14.9 kilometers of induced polarization, resistivity surveys, and magnetic surveys. Since November 1995, the property has been subject to extensive geological, geochemical and geophysical surveys and exploratory drilling to identify drilling targets to define a mineralized gold and silver system. Drilling is discussed in Section 11 of this report. Additionally, surface and underground sampling has exceeded 10,000 rock-chip samples.

This exploration has defined a series of mineralized veins and stockworks which form the basis for the Dolores Mine.

No assays determined prior to Minefinders’ tenure on the property have been used in preparation of their report.

Exploration drilling at Dolores has been oriented toward expanding the resource below and peripheral to the open pit mine.

31  

 




10.0 DRILLING

Drilling prior to 2008 was adequately described in Gustavson (2008), to which report the reader is referred. Drilling and logging procedures have not changed materially since 2007.

The additional in-resource drilling totals 125 holes (2008 resource used 803 holes of 896 existent at the time). Of the additional 125 holes, 29 holes had been drilled and sampled and 64 other holes were geotechnical, excluded, or completely out of the resource area. The 2011 resource is based on 928 holes (992 holes less the 64 which remain immaterial).

32  

 




11.0 SAMPLE PREPARATION, ANALYSES AND SECURITY

No changes to the sample preparation, analysis procedures or security have occurred since the previous Technical Reports (CAM, 2007; Gustavson, 2008). Additionally no new assay data is available that would materially affect the resource estimate done by Minefinders and audited by CAM. The data prepared and presented by Minefinders were audited by at least three other independent engineering firms.

11.1 QA/QC on Drilling Since Last 43-101

The additional in-resource drilling totals 125 holes (2008 resource used 803 holes of 896 existent at the time). Of the additional 125 holes, 29 holes had been drilled and sampled and 64 other holes were geotechnical, excluded, or completely out of the resource area. The 2011 resource is based on 928 holes (992 holes less the 64 which remain immaterial). Hence 13.5% of the drilling is material and not previously considered.

There are no major changes in the QA/QC routines, and no changes in the drilling or sampling routine for the new drillholes. However, there has been no formal internal report generated to date which covers the QA/QC on the new drilling. There have been checks made to duplicates, blanks, and standards during data review, but there has as yet been no formal report generated covering this period.

CAM requested but did not receive the QA/ QC data on the new drilling. However, based on previous work by Minefinders and the fact that the new drilling is less than the 15% of the total material drilling used in the resource estimate, CAM believes it is acceptable to use this data without independent review.

33  

 




12.0 DATA VERIFICATION

The sample database used for the resource has been maintained by Minefinders Exploration group using industry – standard GEMS software. Assay data are acquired and loaded electronically, while geologic information is hand-keyed into spreadsheets prior to loading.

CAM uses automated data processing procedures as much as possible in constructing and auditing geologic databases to assure consistency and minimize errors and costs. These procedures depend heavily on consistent alphanumeric attribute codes and consistent and non-duplicated field labels and drillhole IDs. While many of the issues flagged by these automated procedures are obvious to a human, CAM requires a clean and consistent database before proceeding with geological modeling. Common inconsistencies include:

1.     

Misspellings.

2.     

Confusion of 0 (zero) and O or o.

3.     

Inconsistent use of upper and lower case.

4.     

Inconsistent usage or space _ and -.

5.     

Trailing, leading or internal blanks.

6.     

Inconsistent use of leading zeros in hole IDs.

7.     

Inconsistent analytical units (e.g. PPM, PPB, opt, %, etc.).

8.     

Inconsistent coordinate systems and units (e.g. NAD27 and state plane and mine grid: feet and meters.

For manually generated databases, CAM generally regards an error rate of less than one in 500 good, an error rate of less than one in 100 acceptable and an error rate greater than two in 100 as unacceptable. The acceptability or unacceptability of the database also depends heavily on the impact of the errors. Hence the values for acceptability in unacceptability may easily change by an order of magnitude depending on the nature of the errors. For example a dropped decimal point in a value of 37 for an actual value is 0.37 is much more serious than the entry of a 0.36 for a 0.37. For computer-generated databases any errors may be indicative of problems in data processing procedures and these require resolution at the source of the problem.

The CAM check procedure generates a number of false positives (possible issue which are actually correct). In general if the number of items flagged is less than 2% of the total records the database is acceptable.

CAM also reviews the procedures used to prepare the database and is particularly critical of the common practice of cutting and pasting to obtain the database.

34  

 




Different companies and even geologists within the same company have different methods for drilling, sampling, sample prep and analysis and record-keeping. In some cases it may be necessary to de-weight the results of certain drilling campaigns or types of drilling (e.g. closely-spaced drilling around an initially-successful hole, in order to provide metallurgical samples, or for geostatistical tests).

Over the years CAM personnel have developed a procedure for mathematical and statistically validating exploration databases. This check procedure includes:

  • Check for duplicate collars.

  • Check for twin holes.

  • Check of surface collared holes against surface topography.

  • Check for statistically anomalous downhole surveys.

  • Calculate approximate difference in XYZ location due to differences in hole desurvey algorithms.

  • Check for overlapping assays.

  • Check for zero length assays.

  • Check for long assay intervals.

  • Review of assay statistics by grade class.

  • Review of assay statistics by length class.

  • Checks for holes bottomed in ore.

  • Check for assay values successively the same.

  • Check for assay spikes.

  • Check for downhole contamination by decay analysis.

  • Check of total grade thickness in total and by mineral zone.

  • Bias testing between drilling campaigns and drilling type as appropriate.

In evaluating an existing database CAM uses values flagged by these automated procedures as a starting point for database review and has found that if the error rates in the statistically anomalous values is acceptable then the entire database is generally acceptable.

CAM ran its standard check procedures on the database and did find both errors and possible errors. These were provided to Minefinders and although none of these errors or possible errors would have an overall impact on the global resource they might create problems locally. CAM therefore strongly recommends that Minefinders adopt a single master database and correct the errors found by CAM.

Details of the assay databases provided are given in Section 17 (Resources).

35  

 




On the basis of these statistical checks, and the checks done in the past, CAM believes that the Dolores exploration database has been prepared according to industry norms and is suitable for the development of geological and grade models.

36  

 




13.0 MINERAL PROCESSING AND METALLURGICAL TESTING

Mineral processing at Dolores is discussed in Section 19 of this report.

37  

 




14.0 MINERAL RESOURCE ESTIMATES

The resource section was primarily prepared by Thomas Matthews of Minefinders. His original document has been edited for consistency. Specific CAM review items are indicated by mention of CAM in the relevant sentence or paragraph.

14.1 Basis of Estimation

Proven and probable reserves have been estimated as of December 31, 2010. Minefinders personnel prepared these estimates, with independent review by Robert L Sandefur, P.E. of Chlumsky, Armburst and Meyer, LLC. Mr. Sandefur is the Qualified Person as defined by NI 43-101 for this mineral reserve statement.

The mineral Resources and Reserves in this estimate were calculated using the Canadian Institute of Mining, Metallurgy and Petroleum (CIM), CIM Standards on Mineral Resources and Reserves, Definitions and Guidelines prepared by the CIM Standing Committee on Reserve Definitions and adopted by CIM Council on Dec 11, 2005.

In the past, the Dolores model had been estimated within hand-drawn high grade and low-grade envelopes. Separation of the model into grade zones ensures that the high-grade composites are not unduly projected into low-grade zones, and that high grade zones are not overly smoothed with low-grade data. The current model dispenses with the time-consuming manual sectional envelope construction process and uses a probabilistic approach for differentiating high grade from low grade zones for estimation.

Since production was initiated in 2008, Minefinders has used the common mining practice of determining which portions of the deposit are ore versus waste, based on the assays from vertical blast holes. Given the sub-vertical nature of the ore structures, it is not surprising that methodology did not yield entirely satisfactory results. Hence, in October 2010 an RC drilling program for grade control was initiated and fully implemented on December 1, 2010. In addition to the change from vertical blast holes to inclined RC holes for ore control, the mining methods were changed to mine perpendicular to the sub-vertical structures and geologists were assigned to monitor actual mining operations. CAM is in complete agreement with these changes.

In 2010, Minefinders retained AMEC to develop a new resource model using geostatistical techniques for the sub-vertical ore zones are defined by indicators based on grade and then kriging is applied inside the ore zones to estimate the grade of the blocks. This method has a large number of parameters including:

38  

 




1.     

Values selected for assay capping.

2.     

Composite length.

3.     

Indicator thresholds.

4.     

Search radii and orientation.

5.     

Grade estimation methodology.

6.     

High grade restriction

This type of methodology is becoming more and more common for resource estimation, particularly if actual production data are available. However, having so many parameters may make it possible to tune the model to actual production very closely and this may generate a false sense of accuracy with the model. Also, there is risk associated with holes bottomed in ore and the fact that usually the lower portions of the orebody are not drilled as densely as the upper part.

The use of actual production data to calibrate the model is complicated by the fact that Minefinders has recently made significant changes in the mining method and in ore control drilling and sampling. These changes, with which CAM is in complete agreement, better allow for the fact that the ore occurs in high grade sub-vertical structures which trend to the northwest.

The probabilistic resource model represents a significant change in the methodology used for the resource estimate at Dolores. In the past, the sub-vertical structures which contain ore grade material were manually interpreted and reviewed. This prior methodology, while accepted engineering practice, is both time-consuming to develop initially and time-consuming to update. This type of model gives an undiluted resource with no ore loss or dilution. Therefore, the appropriate ore loss and dilution factors must be applied to this type of resource model prior to its conversion to a reserve.

14.2 Mineral Resources

 

14.2.1 Geologic Interpretation

Structural Domains

Solids models were constructed to represent each of 6 structural domains for the resource model. These domains are intended to represent areas of distinctive structural geometry, wherein the primary structural trends which control mineralization are considered to be consistent and separable.

There are 6 domains representing as follows:

39  

 




  • 101, west fault, north half. (North of Chabacan Canyon) This is the western bounding fault of the structural zone, which generally dips –60° to –70° to the west. It is relatively constrained, and usually contains a single main structure.

  • 102, Main structural zone, north half of deposit. Also used for east dike and mineralization east of eastern bounding fault. This zone comprises a series of structures, all averaging –80° west dipping, which constitute the bulk of mineralization north of Chabacan canyon. The same zone is used for estimation of peripheral mineralization east of the main structural zones due to shared dip orientation.

  • 103, west fault, south half. The western bounding fault south of Chabacan canyon tends to be more discrete, slightly steeper (-75°) dipping, and higher grade than north of Chabacan canyon, so it is assigned a separate zone.

  • 104, transverse shear zone between west fault 103 and eastern bounding fault 105, this shear zone comprises a series of transverse structures modeled north and south of the breccia pipe area, with various steeply east-dipping orientations. These structures are individually somewhat more discontinuous than the main west-dipping structures, but they form a definable structural domain in aggregate.

  • 105, Eastern bounding fault south of Chabacan canyon. This is the main eastern fault, the fault tends to be very strong, well delineated, and carries high grades. It averages –75° west dip.

  • 106, Breccia pipe zone. This domain defines an area where explosive brecciation crosscuts the mineralization of zones 104 and 105, forming a fairly continuous sub-vertical high-grade mineralized zone.

Structural Domains, Sectional Views

Cross sections showing structural domains are illustrated in Figures 14-1 through 14-9. The domains are shown in the following colors:

Table 14-1
Identification of Domains
Domain Color Area
101 Magenta west fault, north half
102 Violet main zone, north half
103 Blue west fault, south half
104 Cyan transverse shear zone
105 Yellow east bounding fault
106 Orange breccia pipe zone

 

40  




Figure 14-1 Figure 14-2
Cross Section 1700 Cross Section 2100
Domain 101 at left, Domain 102 at right. Domain 103 at left, domain 102 at right.

 

Figure 14-3 Figure 14-4
Cross Section 2200 Cross Section 2275
Left to right: Domain 103, 104, 105, 102. Left to Right: Domains 103, 104, 105, 102.
East dike is isolated Domain 102 material at right. Notice transverse shear zones in domain 104.

 

41  




Figure 14-5 Figure 14-6
Cross Section 2425 Cross Section 2625
Left to Right: Domains 103, 104, 105, 102. Left to right: Domains 103, 106, 105 (yellow), 102.
Notice transverse shear zones visible in domain 104.  

 

Figure 14-7 Figure 14-8
Cross Section 2775 Cross Section 2800
Domains: Left -103, Center - 104. Right - 105. Domains: Left -103, Center - 104. Right - 105.

 

Figure 14-9
Cross Section 3075
Domains: Left -103, Center - 104. Right - 105.

 

42  




14.2.2 Sample Databases

The sample database used for the resource has been maintained by Minefinders Exploration group using industry – standard GEMS software. Assay data are acquired and loaded electronically, while geologic information is hand-keyed into spreadsheets prior to loading.

Only drillhole information was used for estimation of block grades. There exist electronic databases of underground and surface geochemistry data, but they were not used for calculation of grade into blocks.

The drill database comprises information from approximately 403 RC and 597 Core holes. Additional holes had been drilled and sampled during 2010, but assay results were not available at the time of preparation of the resource estimate.

14.2.3 Databases Provided to CAM

CAM was provided with the MineSight directory for the project (which contains the exploration database) and a Microsoft access file of the exploration database. There were some differences between these two databases as shown in Tables 14-2 and 14-3, respectively.

Table 14-2
Minefinders Dolores 2011 Update
Drilling Statistics from Assay MEDS 207 Database
Item Number Length (m)
Holes 992 255,632.9
Holes with non-collar downhole surveys 870 235,095.7
Non-collar survey records 8,853 216,845.4
Downhole surveys up 176 3,697.1
Downhole surveys down 9,669 213,148.3
Assay intervals (Au207) 144,997 255,632.9
Assayed intervals (Au207) 143,257 244,610.5

 

Table 14-3
Minefinders Dolores 2011 Update
MF11 MDB
Drilling Statistics from Assay Database
Item Number Length (m)
Holes 1,000 258,453.9
Holes with non-collar downhole surveys 879 238,181.6
Non-collar survey records 3,831 223,282.9
Downhole surveys up 92 3,274.2
Downhole surveys down 4,739 220,008.7
Assay intervals (AuA) 140,157 251,292.5
Assayed intervals (AuA) 140,157 251,292.5

 

43  




As discussed below, resource estimates derived from the two databases are equivalent in terms of contained ounces. However the Microsoft Access database had more holes and fewer inconsistencies than the MineSight database, so CAM selected the Microsoft database for use in validation of the model. The Access database also contained 2-meter composites which CAM elected to use for its independent checks of the models. While the difference between the two databases is not substantive in terms of global results, CAM recommends that Minefinders resolve the differences between the two databases.

There is also inconsistency in the units for gold with some assays being reported in ppb and others being reported in ppm. To avoid the possibility of a local blunder, either high or low, one set of units should be adopted.

To validate and check the resource model CAM used the 2-meter composites provided with the Microsoft Access database. Basic statistics on this database are given in Table 14-4.

Table 14-4
Minefinders Dolores 2011 Update
MF11 MDB
Drilling Statistics from 2m composite MDB Database
Item Number Length (m)
Holes 1,000 258,706.2
Holes with non-collar downhole surveys 879 238,181.6
Non-collar survey records 3,831 223,282.9
Downhole surveys up 92 3,274.2
Downhole surveys down 4,739 220,008.7
Assay intervals (AuC2) 128,248 255,613.2
Assayed intervals (AuC2) 128,248 255,613.2

 

14.2.4 Grade Estimation

Capping of High Grades

Gold and silver assay grades are capped at different levels for each domain prior to compositing. This technique is felt to be more representative than capping composites, as it reduces the impact of single isolated very high-grade assays, without unduly limiting the impact of large numbers of consecutive high grade samples.

Capping values were set for each domain after consideration of several capping methodologies, as shown in Table 14-5.

44  




Table 14-5
Capping Values
Domain Au Cap (ppm) Ag Cap (ppm)
101 20 500
102 22 800
103 18 825
104 15 850
105 12 1000
106 22 900

An additional range restricted capping of composites was employed during interpolation, which will be discussed in further detail below.

Compositing

Drill data are composited in 5-meter intervals, starting from the top of the hole. Composites comprised entirely of void space are not assigned a grade (null value).

This is a departure from earlier resource estimates for Dolores, which used 2-meter composite intervals. Use of longer intervals reduces local grade variability and emphasizes the importance of broader zones of mineralization while de-emphasizing narrow mineralized stringers.

Gold Equivalence

Gold equivalency is calculated for convenience in understanding the relative value of different portions of the mineralized system as well as to allow direct comparisons between different generations of resource & reserve data. It is planned to continue to use an internal 75:1 gold equivalency value ratio to maintain the integrity of these comparisons.

This is roughly equivalent to the following recovery-weighted ratio:

$1200 / Ounce Au * 75% Au Recovery = 75:1
$24 / Ounce Ag   50% Ag Recovery  

These values are very similar to the gold and silver values being used to determine value of ore blocks at Dolores presently.

45  

 




Probabilistic Models

In the past, the Dolores model had been estimated within hand-drawn high grade and low grade envelopes. Separation of the model into grade zones ensures that the high-grade composites are not over projected into low-grade zones, and that high grade zones are not over smoothed with low-grade data. The current model dispenses with the time-consuming manual sectional envelope construction process and uses a probabilistic approach for differentiating high grade from low grade zones for estimation.

All composites above an indicator value are assigned a value of 1 and all composites below the value are assigned a 0. These indicators are then kriged. The resulting value gives the probability that the block should fall within the described zone.

For the low-grade zone, a probability threshold of 0.150 (AuEQ) gpt was used. This value was selected because it is slightly lower than the anticipated cut-off grade of the deposit, and because there is a natural break in the data between unmineralized material and slightly mineralized material. The threshold of this break varies in different portions of the system, but is typically in the 0.100 to 0.200 gpt AuEQ range.

For the high-grade zone, a probability threshold of1.0 gpt (AuEQ) was used. This value is selected because it corresponds to a second natural break in the data.

Each block has a low-grade zone probability and a high-grade zone probability calculated by kriging the low-grade and high-grade composite indicators. These probabilities are backloaded to the composites (the probability value for the block at the center of each composite is loaded to the composite data table.)

Blocks with a low-grade zone probability of greater than 0.50 (50%) are considered part of the low grade zone. Blocks with a high-grade zone probability of greater than 0.38 (38%) are considered part of the high grade zone. This threshold was selected for the high-grade zone by comparison with a nearest neighbor interpolation of high-grade indicator composites.

Selection of 0.38 Probability Threshold for High-Grade Zone

The selection of 0.38 probability threshold for high-grade zone is described below.

High-Grade Indicator Shell Threshold Selection

Selection of the indicator threshold for the high-grade zone turns out to be an issue of fairly high sensitivity, as it has a pronounced impact on the tonnage of the high-grade zone, as well as the mean grade of the deposit. For areas proximal to drilling, a simple 0.50 threshold might be sufficient, but

46  




because of the number of samples used in the interpolation, as one moves away from data, the high-grade zone tends to pinch out, only to reappear as one moves close to another drillhole containing high-grade values. This results in an understatement of the high-grade zone volume, particularly further away from drill data.

In order to address this condition, it was decided to adjust the indicator threshold, and that two tests should be employed to check the extent of volumetric bias being introduced.

Indicator Shell Testing, NN Grade Comparison

The first test employed was to compare average nearest neighbor grade estimated using shells at various threshold values (0.45, 0.40, and 0.38) with nearest neighbor grade estimated without a high-grade shell. The expectation is that in an unbiased condition, the NN average grade for each of these estimates should be identical. If it is not identical, the implication is that the envelope at that threshold is causing a bias in grade.

Table 14-6 shows a comparison of nearest neighbor hard boundaries models vs. soft boundaries models at 0.45, 0.40, 0.38 thresholds. The block counts and mean grades are for high-grade zone and low-grade zone combined. Notice that the soft boundaries models are expected to be the same with different thresholds (compare NNSHC3 with NNSH38), (the difference between NNSH45 and the others is due to the change in capping methodology). The mean grade shift for the NN model is much reduced at the 0.38 threshold.

Table 14-6
Threshold Check - NNHH vs. NNSH (NN Threshold vs. No Threshold)
    Blocks Au Ag Au Ag
Type Description   Mean Mean % %
    (n) (ppm) (ppm) Change Change
0.45 Threshold (Oct 2009 Capping)        
Hard Boundary NNHH45 448,965 0.446 25.032 -  
Soft Boundary NNSH45 448,965 0.485 27.198 8.7% 8.7%
0.40 Threshold (Dec 2009 Capping)        
Hard Boundary NNHHC3 449,995 0.479 26.456 -  
Soft Boundary NNSHC3 450,234 0.496 27.590 3.7% 4.3%
0.38 Threshold (Dec 2009 Capping)        
Hard Boundary NNHH38 450,525 0.489 26.982 -  
Soft Boundary NNSH38 450,807 0.497 27.620 1.6% 2.4%

Based on this test, it was determined that the 0.38 threshold is the least biased of these shells, with only a 2% conservative distortion in estimated grade.

47  




Indicator Shell Testing, NN

A similar test for volume was also conducted, wherein the volume of each indicator envelope was compared with the indicator volume generated by nearest neighbor extrapolation of composites inside the volume. Thus, this is a test of whether the volume of the indicator shell is consistent with the declustered volume represented by the composites within the shell.

 
Figure 14-10
NN Volume Check for 0.38 High-Grade Indicator Threshold

48  





Figure 14-11
NN Volume Check for 0.40 High-Grade Indicator Threshold

It can be observed from these graphs that the volumes represented match very closely for low distances from drillholes, but that they increase with distance from data. This is due to the increased smoothing with increasing distance. The target here is to have a slightly conservative comparison at distances approximating the M&I distance search (35m).

The 0.38 threshold shows better agreement in volumes throughout the range of distances from data. The volume of the high-grade zone will still be somewhat understated at increasing distances in the inferred material, but additional drilling within the inferred material will normally correct this impact.

Typical sections showing Low-Grade and High-Grade zone geometry are shown in Figures 14-12 through 14-14.

49  





Figure 14-12
Cross Section 1925

 
Figure 14-13
Cross Section 2675

50  




 
Figure 14-14
Cross Section 2225

Boundary Conditions between High-Grade and Low-Grade Zones

A ‘firm’ boundary is used to handle data across the boundary between the high-grade and low-grade zones. Use of a ‘hard’ boundary (high-grade zone composites only impact high-grade zone blocks and low-grade zone composites only impact low-grade zone blocks) with kriging gives a distortion in the grade-tonnage curves because low-grade zone blocks tend to be assigned grades similar to the average grade of the low-grade zone composites and high-grade zone blocks similarly tend to be averaged toward the average grade of the high-grade zone composites. This understates the higher-grade portion of the low-grade zones and understates the lower-grade portion of the high-grade zone, leaving a ‘valley’ in the grade distribution at the 0.9-1.1 gpt values.

Conversely, it was found that if a ‘soft’ boundary condition is used (i.e., allowing full exchange of information between the high-grade and low-grade zones), (equivalent to not creating a high-grade and low-grade zone), there tends to be an averaging all the data in both high-grade and low-grade zones toward the mean composite values for both zones. This increases the variability of block grades relative to composite grades locally, which gives excess blocks in the middle of the grade distribution, and understates the number of blocks at the high end and low ends of the distribution.

The ‘firm’ boundary allows composites adjacent to the boundary between the high-grade and low-grade zones to impact computation of block grades in the adjacent zone. Thus, a high-grade composite at the high-grade/low-grade boundary would be able to influence grade both in the high-grade zone and in the

51  




adjacent low-grade zone. Similarly, low-grade composites at the boundary are able to influence the grade of the adjacent high-grade domain.

The probability value stored within each composite is used to flag the composites as either 1, low-grade only, 2, low-grade adjacent to high-grade, 3, high-grade adjacent to low-grade or 4 high-grade only. Composites flagged as 1, 2, 3 are used for estimation of grades within low-grade zones. Composites flagged as 2, 3, 4 are used for estimation of grades within high-grade zones.

Range Restricted Capping

A second capping methodology is applied to the interpolation during the kriging process. In order to constrain the potential impact of high-grade ‘3’ composites on the low-grade zone, as well as to constrain the impact of high outlier low-grade zone composites, a range restricted cap is applied to the composites.

Blocks within 20 meters of the composite use the (already capped) composite grades for estimation. Blocks beyond 20 meters from the composite use the threshold value for that composite in the estimation. This has the effect of reducing over-projection of higher grade gold and silver values into the low-grade zone.

Similar range-restricted capping for gold is applied within the high-grade zone, as there is lower observable continuity in the high-grade gold as compared to high-grade silver. In the high-grade zone, a 30 meter range restriction is used for gold composites, with caps as shown in Table 14-7.

Table 14-7
Range Restricted Caps
  Low-Grade Low-Grade Low-Grade High-Grade High-Grade High-Grade
Domain  Range Gold Silver Range  Gold Silver
    Threshold Threshold   Threshold Threshold
101 20m 2.0 100 30m 8.0 N/A
102 20m 5.0 250 30m 15.0 N/A
103 20m 3.0 100 30m 9.0 N/A
104 20m 1.0 150 30m 5.0 N/A
105 20m 1.5 250 30m 7.0 N/A
106 20m 3.0 150 30m 10.0 N/A

Geostatistics

Correlograms were constructed for each structural domain and for both gold and silver composites. Ranges and sills from these correlograms are used as the kriging parameters for the estimation. Gold

52  




variograms by domain are illustrated in Figures 14-15 through 14-20. Silver variograms by domain are illustrated in Figures 14-21 through 14-26.

53  




Figure 14-15 Figure 14-16
Domain 101 – Au Variogram Domain 102 – Au Variogram

Figure 14-17 Figure 14-18
Domain 103 – Au Variogram Domain 104 – Au Variogram

Figure 14-19 Figure 14-20
Domain 105 – Au Variogram Domain 106 – Au Variogram

 

54  




Figure 14-21 Figure 14-22
Domain 101 – Ag Variogram Domain 102 – Ag Variogram

Figure 14-23 Figure 14-24
Domain 103 – Ag Variogram Domain 104 – Ag Variogram

Figure 14-25 Figure 14-26
Domain 105 – Ag Variogram Domain 106 – Ag Variogram

 

55  




14.2.5 Block Model

Block model dimensions for this resource estimation are the same as those used for the ore control system. Blocks are 3-meters across strike, 5-meters along strike and 7.5-meters tall. The block model is oriented with the long axis parallel to the 330° primary strike of the deposit. The maximum model elevation is 1800 meters (block crest) and the minimum elevation is 997.5m, (block toe).

The block model contains 600 columns, 700 rows and 107 levels.

Models

The model last reported in a 43-101 was based on constraining grade to the sub-vertical structures interpreted by Minefinders. A model of this type conforms to accepted engineering practice, but its construction is very time-consuming and labor-intensive. For this reason in 2010, Minefinders contracted with AMEC to develop a model based on geostatistical probabilistic methodology. This model can be updated very rapidly, but is not without some risks as discussed above.

Geometric parameters of the probabilistic model are given in Table 14-8.

Table 14-8
Minefinders
Dolores 2011 PROBABILISTIC Model
Geometric Parameters
Origin (Meters) Number of Block Size (Meters)
Northing 50597.00 Rows 700 Row 5.0
Easting 50336.00 Columns 600 Column 3.0
Elevation 997.5 Benches 107 Bench 7.5
Rotation Angle (330.00)

This probabilistic model covers the same area as the constrained inverse distance squared model, parameters for which are given in Table 14-9.

Table 14-9
Dolores 2011 Constrained ID2 Model
Geometric Parameters
Origin (Meters) Number of Block Size (Meters)
Northing 50597.00 Rows 350 Row 10.00
Easting 50336.00 Columns 600 Column 3.00
Elevation 1000.00 Benches 160 Bench 5.00
Rotation Angle (330.00)

 

56  

 




If Minefinders elects to continue to use the constrained ID2 model as a check, then CAM recommends that the model geometric parameters be changed to be the same as the probabilistic model.

CAM performed independent calculations of the resource using several different techniques; two of which were nearest neighbor estimates. Because of the decision to use the 2-meter composites, it is best practice to have the model block size smaller than the composite length. The CAM model which covers the same area as the other two Minefinders’ models would therefore have a significantly smaller block size for columns and benches. Because most of the drilling was perpendicular to the strike of the orebody, a row size of 5-meters (the same as in the probabilistic model) was used.

The geometric parameters of the CAM check model are listed in Table 14-10.

Table 14-10
Dolores 2011 CAM CHECK MODEL
Geometric Parameters
Origin (Meters) Number of Block Size (Meters)
Northing 50597.00 Rows 700 Row 5.00
Easting 50336.00 Columns 1800 Column 1.00
Elevation 1000.00 Benches 800 Bench 1.00
Rotation Angle (330.00)

Grade Estimation Methodology

Grades are estimated using MineSight software. Grade interpolation is done separately for gold and silver. Interpolation is done by ordinary kriging using different search ellipses and nugget values for gold and for silver, as well as for each structural domain, for a total of 24 interpolation passes.

Interpolation uses a minimum of 3 and a maximum of 8 composites for estimation. A sample interpolation profile is listed in Table 14-11

Table 14-11
Ranges and Rotations for Kriging Interpolation, High-Grade Zone Gold
Domain 101, Au
High-
grade
102, Au
High-
grade
103, Au
High-
grade
104, Au
High-
grade
105, Au
High-
grade
106, Au
High-
grade
Min Comps 3 3 3 3 3 3
Max Comps 8 8 8 8 8 8
Max Comps / Hole 2 2 2 2 2 2
Comp Length 5m 5m 5m 5m 5m 5m
             
Search Y 150 200 250 150 160 200
Search X 20 50 25 50 20 70

 

57  

 




Table 14-11
Ranges and Rotations for Kriging Interpolation, High-Grade Zone Gold
Domain 101, Au
High-
grade
102, Au
High-
grade
103, Au
High-
grade
104, Au
High-
grade
105, Au
High-
grade
106, Au
High-
grade
Search Z 120 100 130 100 100 150
             
Nugget .403 .402 .412 .410 .421 .394
             
Primary Model Spherical Spherical Spherical Spherical Spherical Spherical
Rotational System GSLib GSLib GSLib GSLib GSLib GSLib
Primary Sill .472 .462 .312 .480 .526 .318
Prim Rot 1 -30 11 -10 -23 -30 49
Prim Rot 2 0 75 -27 28 0 68
Prim Rot 3 65 -38 -35 -62 70 -70
Prim Maj Range 26 18.1 16.1 32.1 28.7 16.8
Prim Min range 18 8 6.2 10.1 17.3 18.4
Prim Vert range 10.7 21.3 16.1 4.3 12.1 18.1
             
Secondary Model Spherical Spherical Spherical Spherical Spherical Spherical
Secondary Sill .125 .136 .276 .110 .053 .288
Sec Rot 1 -30 11 -10 -23 -30 49
Sec Rot 2 0 75 -27 28 0 68
Sec Rot 3 65 -38 -35 -62 70 -70
Sec Maj Range 110.7 149.4 35.3 164 155 150
Sec Min Range 39.3 43.5 22.3 94.7 142.8 57.4
Sec Vert Range 60 274.3 69.3 54.7 63.8 241.4

 

14.2.6 Tonnage Estimation

Resource Definitions

Bulk Density

Bulk density for all blocks is defaulted at 2.55 t/m3. This is a weighted average of bulk density measurements for andesite (2.57 t/m3) and Latite (2.53 t/m3). There is substantial complexity in the geometry of the latite dike swarm at Dolores and it has not been considered practical or necessary to create solids for density tagging of the material types given the low variability of rock density.

58  




Mined out Volumes

Block grades are estimated for all blocks below original mine topography. Resource and reserves reporting reports only include those blocks below the January 1 2011 surveyed topography surface from the mine.

Underground workings have been encountered both in drilling and in mining in benches in phase 2 and phase 3 of the open pit. There has been no allowance made to remove tonnes and grade associated with the old underground workings. It is expected that the tonnages contained within these workings are very small as compared to the scale of the resource model. The workings are typically less than one block in width, and the exact locations are not fully understood. There has been no observable impact on reconciliation for pit phases where mine workings are present in the benches.

Classification

Resource classification is based primarily on distance from data. A simplified interpolation ellipse was used, consistent with the scale of distance search used in the 2008 resource model, to flag blocks with distance to the closest composite.

Blocks within a 15 x 20 x 17.5 search from composites are counted as measured. Blocks within a 30 x 40 x 35 search of composites are counted as indicated. More distal estimated blocks are counted as inferred. This search ellipse is exactly identical to the ellipse used in the 2008 model.

The 2008 resource model employed drill data as well as surface and underground data for distance classification. The present model uses only exploration drill data. In this respect, the 2011 model is more conservative, as it uses only drill data for classification of resources.

Cut-Off Grade & reporting

Resource reporting is done at a series of gold-equivalent (75:1 ratio) cutoffs which roughly correspond to a range of probable cut-off grades. In actual practice the mine uses a recoverable dollar value cutoff to account for differences in recoveries for gold and silver, which vary according to alteration type. The current operating cutoff at the mine is $7.50 per tonne, although the best available material is always sent through the crusher first, which raises the nominal cutoff to $10 most months, with the lower grade material being stockpiled. This corresponds roughly to 0.3 gpt AuEQ grade.

59  




14.2.7 Reconciliation

Model Validation by CAM

To validate the total resource, CAM calculated contained tonnes, grade and contained ounces using a gold to silver ratio of 75 to 1 and reporting resources between original surface topography and a pit surface provided by Minefinders based on measured, indicated and inferred. Because of regulatory restrictions on reporting totals of measured, indicated and inferred it is necessary to report these numbers as a percentage of the base case MineSight probabilistic model. These results are presented in Table 14-12

Table 14-12
Percent of MineSight
Model % of
Tonnes
% of
Grade
% of
Ounces
MineSight Probabilistic 100.0 100.0 100.0
Constrained ID2 82.1 124.3 102.1
CAM NN 100x100x3 71.9 145.9 104.9
CAM NN 100x100x100 72.7 148.1 107.7
CAM Sector 100x100x3 6 max 87.2 112.4 98.0
CAM Sector 100x100x3 12 max 93.4 102.8 96.1
CAM Sector 200x200x6 6 max 87.2 112.1 97.8
CAM Sector 200x200x6 12 max 93.3 102.5 95.6

The CAM sector searches of 100 x 100 x 3 and 200 x 200 x 6 have the long dimensions along strike (azimuth of 330°) and vertical. The short dimension is perpendicular to the average strike of the ore zones. The sectors are oriented along the six faces of a rectangular solid containing the search ellipsoid.

All of the models, except one, check within the plus or minus 5% which CAM regards as good. The exception is the CAM nearest neighbor check with the 100-meter isotropic search in terms of total ounces. CAM is not concerned about this 7.7% difference for the unconstrained nearest neighbor model as it predicts more ounces and is geologically unreasonable given the geometry of the deposit.

Since all of the models are equivalent in terms of total contained ounces, within the usual standards of reproducibility between models, the decision on which model to use should be what based on how well the model matches actual production.

One of the risks associated with a non-geologically constrained model is that higher grade composites at or near the bottom of a hole can extrapolate ounces into regions where there is insufficient supporting data. To verify this was not occurring with the MineSight probabilistic model, CAM constructed swath

60  

 




plots showing ounces per meter across strike for both the MineSight probabilistic model and the constrained ID2 model as shown in Figures 14-27 and 14-28.

61  

 





62  

 




Both figures show about the same across-strike distance, indicating that over the entire length of the orebody there is not a problem with high grade at the bottom of the whole being excessively extrapolated. However, it is possible that there may be local issues and CAM recommends that Minefinders calculate ounces per hole and ounces per composite to see if there are any anomalous values which require local attention.

Production Data

Since production was initiated in 2008, Minefinders has used the common mining practice of determining which portions of the deposit are ore and waste based on the assays from vertical blast holes. Given the sub-vertical nature of the ore structures it is not surprising that methodology did not give entirely satisfactory results. Hence, in October 2010 an RC drilling program for grade control was initiated and fully implemented on December 1, 2010. In addition to the change from vertical blast holes to inclined RC holes for ore control, the mining methods were changed to mine perpendicular to the sub-vertical structures and geologists were assigned to monitor actual mining operations. CAM is in complete agreement with these changes.

CAM was provided with the duplicate assays on some of the RC samples as well as the QA/QC document for reviewing RC duplicates. The criteria used for acceptability is a simple plus or -10% criteria which does not conform to best engineering practice. CAM recommends that this criteria be expanded to include scatter plots of original value versus duplicates on both a untransformed and log-log basis.

Most of the RC grade control holes are 43-meters long and drilled at in azimuth of 60° and a dip of 45° and are sampled on intervals ranging from 1.5 to 2 meters.

The reconciliation provided by Minefinders along with some additional calculations on equivalent grade and contained equivalent ounces are summarized in Table 14-13.

Table 14-13
Reconciliation (Provided by Minefinders)
For December 1,2010 Through March 31,2011 (4 Months)
Item 2011 Model Ore
Polygons
Process` Difference CAM Check
Difference (%) *
Tonnes 2,475,576 2,284,583 2,421,833 -2.17% -2.17
Au Grade (gpt) 0.529 0.499 0.515 -2.60% -2.65
Ag Grade (gpt) 50.7 45.115 49.232 -2.90% -2.90
AuEQ75 (gpt) * 1.205 1.101 1.171   -2.79
AuEQ75 (Contained) * 95,908 80,835 91,212   -4.90
* Calculated by CAM          

63  




To validate the provided reconciliation, CAM constructed a nearest-neighbor model based on the RC assays and then compared the 2-meter exploration composites against this model based on distance to the nearest RC assay. This comparison for maximum distance of 10 and 20 meters from the nearest RC assay is given in Table 14-14.

Table 14-14
RC assays vs. 2m Exploration Composites Bias test
Dist. Au ppm Ag ppm
RC Exp t.test
Prob
RC Exp t.test
Prob
Cnt Mean Cnt Mean Cnt Mean Cnt Mean
10 23906 0.29 1320 0.307 0.76174 23899 25.915 1320 24.444 0.37285
20 23906 0.29 2357 0.308 0.69524 23899 25.915 2357 22.196 0.00584

Because no cutoff is used in Table 14-14, the mean grade for both gold and silver are lower than in Table 14-13.

Only four months of production data with RC grade control fully implemented were available at the data close off date. The Minefinders reconciliation was within 5% for tonnes, silver grade, gold grade and contained equivalent ounces, but it is possible that the 5% reconciliation is due to the model estimation parameters and may not be indicative of future performance. To calculate the statistical uncertainty associated with only having four months of production data, CAM calculated the low 10% confidence limit of the mean grade of 2-meter exploration composites contained within the four months of production as shown in Table 14-15.

Table 14-15
Approximate (~) Statistical Uncertainty in Resource Estimates Based on 2m Composites in 4 months of Pit Progress
Cutoff Count Mean StdDev Coef.
Variance
(CV)
~CV
Grade
~CV
Tonnes
~CV
Ounces
~ Low 0.10
Confidence
Limit
0.00 1233 0.423 1.248 2.948 0.084 0.028 0.089 -0.114
0.20 336 1.431 2.078 1.452 0.079 0.055 0.096 -0.123
0.50 224 1.988 2.355 1.185 0.079 0.067 0.104 -0.133

Assuming a cutoff of 0.5 gram per tonne equivalent gold, the low 10% confidence limit is approximately -13% as opposed to the 3% given in the reconciliation. Although there are a number of statistical approximations involved in these calculations, it appears that at least a year and probably two will be required to determine if the 3% performance is correct or just a result of tuning the estimation parameters.

64  




This long time period is required because halving the statistical uncertainty requires four times as many samples.

14.2.8 Changes in the Resource Estimation Methodology and Impact of Changes

There are several changes in the resource estimation methodology as compared to the 2008 43-101. In general, the changes are intended to decrease the local grade variability of the resource block model so that the resource model more closely reflects the tonnages and grades actually yielded by mining. The changes include:

  1.     

The use of probabilistic rather than hand-constructed LG and HG zone boundaries.

  2.     

Block size: Block size for the 2008 model was 3m x 10m x 5 m (x,y,z). The 2011 models use the 2m x 5m x 7.5m block size used for ore control. There is essentially no difference in the resource due to changed block size.

  3.     

The 2008 resource was reblocked to 6m x 10m x 7.5 m (x,y,z) to simulate mining dilution. 2008 reserves reported the reblocked model. The 2011 model incorporates change of support and dilution in the compositing and block estimation, rather than by reblocking.

  4.     

Composite lengths have been increased from 2m to 5m for the 2011 model. This increases downhole smoothing of grade, and decreases local grade variability.

  5.     

Interpolation methodology: The 2011 model uses ordinary kriging as opposed to the inverse distance estimation employed in the 2008 model. Kriging is more effective in handling variably spaced composite information than is inverse distance, and has the effect of smoothing the distribution of the block grades.

     
14.2.9 Resource Tabulation

Resources tabulations are summarized in Tables 14-16 through 14-19. Note that because of the estimation method, the resource includes diluted ore reserves.

65  

 




Table 14-16
Mineral Resources – Measured
Cut-off
(gpt AuEq)(i)
Tonnes
(million)
Gold Gold Silver Silver Cut-off
Ounces Grade Ounces Grade (gpt
(million) (gpt) (million) (gpt) AuEq)(i)
0.2 96,587,826 1,605,129 0.517 83,718,850 27.0 1.036
0.3 80,050,610 1,530,279 0.595 79,241,979 30.8 1.187
0.4 61,857,757 1,415,775 0.712 72,508,982 36.5 1.414
0.5 46,836,549 1,293,629 0.859 65,502,085 43.5 1.696
0.6 36,734,441 1,194,358 1.011 59,652,930 50.5 1.982
1.0 21,765,502 989,672 1.414 48,087,568 68.7 2.735
1.5 14,558,872 823,659 1.760 39,004,084 83.3 3.362
2.0 9,291,428 652,792 2.185 83,718,850 99.9 4.106
(i) Gold-equivalent (AuEq) mine cut-off grades are based on a silver to gold ratio of 75 to one.
(ii) AuEq grade is based on a silver to gold ratio of 52 to one.

 

Table 14-17
Mineral Resources – Indicated
Cut-off
(gpt AuEq)(i)
Tonnes
(million)
Gold Gold Silver Silver Cut-off
Ounces Grade Ounces Grade (gpt
(million) (gpt) (million) (gpt) AuEq)(i)
0.2 90,584,324 1,203,357 0.413 65,206,084 22.4 0.844
0.3 71,856,401 1,116,743 0.483 60,272,682 26.1 0.985
0.4 52,033,568 988,672 0.591 53,231,017 31.8 1.203
0.5 37,090,620 865,371 0.726 46,434,526 38.9 1.474
0.6 28,057,187 775,032 0.859 41,331,407 45.8 1.740
1.0 14,534,863 589,634 1.262 30,790,123 65.9 2.529
1.5 9,219,244 463,983 1.565 24,506,244 82.7 3.155
2.0 5,671,306 351,683 1.929 18,134,789 99.5 3.842
(i) Gold-equivalent (AuEq) mine cut-off grades are based on a silver to gold ratio of 75 to one.
(ii) AuEq grade is based on a silver to gold ratio of 52 to one.

 

66  

 




Table 14-18
Mineral Resources – Measured + Indicated
Cut-off
(gpt AuEq)(i)
Tonnes
(million)
Gold Gold Silver Silver Cut-off
Ounces Grade Ounces Grade (gpt
(million) (gpt) (million) (gpt) AuEq)(i)
0.2 187,168,418 2,808,351 0.467 148,932,250 24.8 0.934
0.3 151,904,139 2,646,975 0.542 139,478,988 28.6 1.092
0.4 113,889,315 2,404,534 0.657 125,737,334 34.3 1.317
0.5 83,925,445 2,159,102 0.800 111,948,445 41.5 1.598
0.6 64,789,904 1,969,472 0.946 100,983,618 48.5 1.879
1.0 36,299,504 1,579,224 1.353 78,879,494 67.6 2.653
1.5 23,777,543 1,287,636 1.684 63,510,340 83.1 3.282
2.0 14,962,161 1,004,446 2.088 47,963,848 99.7 4.005
(i) Gold-equivalent (AuEq) mine cut-off grades are based on a silver to gold ratio of 75 to one.
(ii) AuEq grade is based on a silver to gold ratio of 52 to one.

 

Table 14-19
Mineral Resources – Inferred
Cut-off
(gpt AuEq)(i)
Tonnes
(million)
Gold Gold Silver Silver Cut-off
Ounces Grade Ounces Grade (gpt
(million) (gpt) (million) (gpt) AuEq)(i)
0.2 40,786,301 388,139 0.296 19,459,389 14.8 0.581
0.3 27,632,397 326,214 0.367 16,133,009 18.2 0.717
0.4 17,359,931 256,959 0.460 12,764,245 22.9 0.900
0.5 11,089,833 202,514 0.568 10,118,552 28.4 1.114
0.6 7,451,842 164,086 0.685 8,219,877 34.3 1.345
1.0 2,755,033 96,750 1.092 4,788,327 54.1 2.132
1.5 1,323,119 62,140 1.461 3,223,125 75.8 2.919
2.0 668,668 40,646 1.891 2,087,209 97.1 3.758
(i) Gold-equivalent (AuEq) mine cut-off grades are based on a silver to gold ratio of 75 to one.
(ii) AuEq grade is based on a silver to gold ratio of 52 to one.

Mineral resources which are not mineral reserves do not have demonstrated economic viability. The estimate of mineral resources may be materially affected by environmental, permitting, legal, title, taxation, socio-political, marketing, or other relevant issues.

The quantity and grade of reported inferred resources in this estimation are conceptual in nature. There has been insufficient exploration to define these resources as indicated, and it is uncertain if further exploration will result in conversion to an indicated or measured mineral resource.

67  

 




15.0 MINERAL RESERVE ESTIMATES

 

15.1 Pit Optimization, Pit Design, and Economic Parameters

A standard methodology for pit limit analysis was followed by Minefinders using the Whittle software package which applies an implementation of the Lerchs-Grossmann (LG) algorithm. The parameters for pit optimization are summarized in Table 15-1.

15.1.1 Pit Optimization Parameters

 

Table 15-1
Pit Optimization Parameters
Item Value
Metal Prices  
Gold ($/troy ounce) 1200
Silver ($/troy ounce) 23
   
Block dimensions
(meters)
X = 3.0
Y = 5.0
Z = 7.5
Reblocking X = 2,
Y = 2
Z = 1
Reblocked Block Dimensions
(meters)
X = 6.0
Y = 10.0
Z = 7.5
   
Pit Slope Profiles Profile 1
Above 1350 elevation
Bearing 60°, slope 48
Bearing 240°, Slope 50.
Profile 2
Below 1350 elevation
Bearing 60°, slope 46
Bearing 240°, Slope 48.
   
Mining Cost Range: $1.10 - $1.53 / tonne
Increasing with depth.
Process Cost (Heap Leach) $4.21 / tonne
G&A Cost $14.4 million / year
$2.47 / tone
   
Heap Leach Production Rate 5,832,000 tonnes per year
(16,200 tonnes per day)
   
Gold Recoveries  
Oxide 79%
Mixed 79%
Sulfide (LG) 56%
Sulfide (HG) 65%
MnOx latite 78%

 

68  

 




Table 15-1
Pit Optimization Parameters
Item Value
   
Silver Recoveries  
Oxide 46%
Mixed 51.5%
Sulfide (LG) 55%
Sulfide (HG) 71%
MnOx latite 21%
   
Selling Costs - Gold $0.71 per ounce
Selling Costs – Silver $0.22 per ounce
   
Refining Recovery – Gold 99.8%
Refining Recovery – Silver 99.75%
   
NSR – Gold 3.25%
NSR – Silver 2.0%
Whittle Optimization 0.3 – 1.50
(increments of 0.015)

Metals price assumptions for the resource estimate are based on a weighted average of the 36-month trailing average metals prices (60%) and the 24-month future price projections (40%), as of February 28, 2011. The gold and silver metal prices calculated by CAM as of February 28, 2011 are shown in Table 15-2.

Table 15-2
Projected Metal Prices as of February 28, 2011 (CAM)
    Projected
Metal Units (36-Mon. Avg @ 60%
    + 24-Mon. Future @ 40%)
Gold USD per troy oz. 1203.71
Silver USD per troy oz. 23.75

Based on the stated prices, metal prices used in pit optimization were set to $1200 for gold and $23 for silver.

69  

 




15.1.2 Pit Shell Optimization

The Whittle LG pit shell selected for the reserves statement is pit 29, corresponding to a revenue factor of 0.720, or $864 gold and $16.56 Silver. Figure 15-1 illustrates the quantities and values of the series of Whittle LG pits.

Figure 15-1
Graph of Whittle LG Pits

Whittle pits shells represent an approximation of the material to be mined, excluding minable access and mining considerations. The planned inter-ramp highwall angle is 50°, but in Whittle the average overall angle is reduced to accommodate ramps, particularly at the bottom of the pit to account for haulage access.

15.1.3 Ultimate Pit with Haul Roads

An ultimate pit plan including haul roads and safety berms was created using Minesight software by the Dolores planning engineering department. Haul roads were designed with maximum 10% grade and 25 meter width. 6.4m wide catch benches were designed every other bench (every 15m), with 67 degree inter-bench face angle. This results in a net inter-ramp highwall angle of 50°, consistent with ‘good’ mining practices and blast control parameters from Golder’s geotechnical recommendations.

70  

 




This ‘ultimate pit’ was used for reporting of reserves.

15.1.4 Production Phases and Dump Location

Mine production phases and dump location plans have been completed by Dolores mine planning engineers covering the ultimate pit volume, but have not been reviewed in detail for the purposes of this reserve update.

15.1.5 Internal Cutoff Calculation

For the purpose of reporting mine reserves, the value of gold and silver in each block, net of recovery are stored (recoveries vary according to alteration for each block). An internal value cutoff grade (excluding mining cost) is used for separation of ore from waste based on the costs shown in Table 15-3.

Table 15-3
Costs Used in Cutoff Grade Calculations
Item Cost per tonne
Leach Processing $4.17
General & Administrative $2.47
Rehandle Allowance $0.25
Pad Construction Allowance $0.61
Total $7.50

Reserves are defined as that material between the January 1, 2011 pit surface and the ultimate pit surface including haul roads, as described above, which supports a $7.50/tonne cutoff.

15.2 Reserve Summary

Minable reserves for the Dolores Mine as of 31-Dec-2010 are summarized in Table 15-4.

Table 15-4
Dolores Mine Reserves as of 31-Dec-2010
Category Tonnes Au (g/t) Ag (g/t) Au Oz Ag Oz
(million) (million)
Probable 62,430,655 0.62 34.2 1.254 68.552
Proven 45,212,669 0.53 31.6 0.770 45.972
Total 107,643,324 0.58 33.1 2.024 114.524

The mineral Resources in this estimate were calculated using the Canadian Institute of Mining, Metallurgy and Petroleum (CIM), CIM Standards on Mineral Resources and Reserves, Definitions and

71  

 




Guidelines prepared by the CIM Standing Committee on Reserve Definitions and adopted by CIM Council on 27 November 2010. Dilution is included, as required by the CIM definition.

The previous public announcement of mineral reserves had been prepared by Gustavson (2008), using the mineral resources prepared by CAM in 2007. The Gustavson figures are shown in Table 15-5, based on material in-place as of 31 December 2007. Table 15-5 shows the sums of sulfide, mixed, and oxide material.

Table 15-5
Dolores Mine Reserves as of 31-Dec-2007
Category Tonnes Au (g/t) Ag (g/t) Au Oz Ag Oz
(million) (million)
Probable 42,675,000 0.72 38.8 0.989 53.229
Proven 56,629,000 0.80 40.3 1.454 73,415
Total 99,305,000 0.77 39.7 2.443 126,644

Compared to year-end 2010, the year-end 2007 reserves show lower tonnages and higher grades, due in part to the higher metals prices in 2010, with resultant lower cutoff grades in the pit shell.

15.3 CAM Review of Reserves

CAM has reviewed the pit optimization parameters, pit design, cost parameters and cutoff calculations and finds them to be reasonable and acceptable. CAM reviewed the ultimate pit design and finds that it conforms acceptably to the Whittle LG pit shell (pit 29) upon which it was based. CAM verified the reserve quantities stated between the 31-Dec-2010 topographic surface and the ultimate pit design.

CAM believes that the mineral reserve estimates are unlikely to be materially affected by unforeseen mining, metallurgical, infrastructure, permitting, or other factors, except those factors which are inherent in all mining project: namely metals prices, unexpected price hikes for fuel or other inputs, civil unrest, or catastrophic natural events.

Minefinders experienced loss of production in 2010 due a tear in the Phase 1 leach pad liner. Excavation and repair work is currently underway on the liner, with completion targeted for Q1, 2012. Cost of this remediation work is estimated at $3.5 milllion. After repair work is completed, Minefinders anticipates that the Phase 1 leach pad will be returned to leach and that recovery of gold and silver contained in the pad will re-commence. The engineering firm Wardell Armstrong has been contracted to oversee ore stacking and operations on phase 2, and to update the design of the phase 3 pad to guard against the possibility of similar failure for additional leach pads.

72  

 




The ore already stacked on the Phase 1 pad is not a portion of the mineral resources or reserves discussed in this report, as the material has already been mined. Further production of gold and silver from the phase 1 pad material would be additional to the reserves discussed in this report.

73  

 




16.0 MINING METHODS

 

16.1 Mining Operations

The Dolores mine provides ore to a crushing plant at a planned rate of 6.48 million tonnes per year. The ore is crushed and conveyed to the leach pad at a rate of 18,000 tonnes per day.

The mine plan has approximately 16.5 years of commercial production remaining. The peak total material movement is 33.7 million tonnes per year, or about 95,500 tonnes per day, (assuming 365 operating days per year). The operation runs 2, 12 hour shifts per day, with the mine labor force working a cycle of 14 days on, 7 days off. This is typical for similar operations in the area.

Overall mining strip ratio for the remaining mine life is 4.47 tonnes mined per tonne of ore, including amounts which may be capitalized as pre-stripping under international financial reporting standards.

16.2 Mine Equipment Fleet

The mine equipment fleet is sized to efficiently mine ore and waste from 7.5 meter bench heights.

Table 16-1
Mine Equipment Fleet
Number Model Description
15 HD785 100 T Haul Truck
2 PC3000 Hydraulic Shovel
1 PC300 Backhoe
2 WA900 Front End Loader
1 WA600 Front Loader (Stockpiles)
2 GD825 Road Grader
1 GD555 Road Grader
2 D375 Dozer
1 D155 Dozer
1 D61 Dozer (Leach Pads)
2 WD500 Wheel Dozer
2 HD785 Water Truck
3 IR DM45 Blasthole Drill
1 ECM780 Blasthole Drill
1 WA470-5 Tire Handler
1 FORD Fuel & Lube Truck
1 FORD Maintenance Welder
2 FORD Maintenance Pickups
1 GROVE 60T Crane
6 Genie Light Plants

 

74  

 




Three additional 100-ton CAT 777 haul trucks entered service in July of 2011. Delivery of 1 RC drilling rig for ore control is expected during August.

16.3 Drill & Blast

Ore and waste are drilled and blasted using 17.1 cm diameter holes with spacing between holes of 4.5 to 5 meters (depending on rock characteristics). Drill and blast in ore areas is single bench, with 3m of burden and 1m subdrill. Drill and blast in waste areas is double benched, using 15m holes on 5.5m grid, 4m of stemming, and 1.5m subdrill. Mining in double-benched areas is carried out in two flitches. Blasting patterns, loads, stemming, et al, are revised periodically as rock conditions and mining require.

Designed pit slope angle is nominally 50°. Pit highwalls are double-benched, with 6.4m wide catch benches designed every other bench (every 15m), with a 67 degreebench face angle. This results in a net inter-ramp highwall angle of 50°, consistent with ‘good’ mining practices and blast control parameters from Golder’s geotechnical recommendations.

16.4 Ore Control

Ore control is currently carried out using angled 42m Reverse Circulation drill holes, with holes drilled without water injection at N60E at -45 degrees, generally perpendicular to the strike and dip of the deposit. Holes are sampled using a gyratory splitter at 2 meter intervals. These holes allow the mine to collect data for 30 meters vertically, allowing for detailed assay information to be available for 4 benches in advance of production.

Ore control assay data is loaded and interpolated daily, and updated mining polygons are supplied to the mine planners on a daily basis. Dig polygon location information is marked on bench by survey crews on a daily basis. Polygon locations are checked and adjusted as necessary by pit geologists, and ore spotters are used to ensure that loader operators honor the boundaries of the dig polygons.

16.5 Mine Personnel

Mine and maintenance personnel are primarily Mexican nationals, with the exception of certain management and technical positions, where expatriates fill certain roles.

Supervision and engineering personnel generally work a 14 day cycle, with 10 days working, and 4 days of rest. Operating personnel generally work a 21 day cycle, with 14 days working and 7 days rest. Transportation is provided to Chihuahua City, Chihuahua, Hermosillo, Sonora, and Madera, Chihuahua. Expatriate personnel generally work 8 weeks on, with 3 weeks rest.

75  

 




17.0 RECOVERY METHODS

Test work has shown gold and silver mineralization in Dolores to be amenable to conventional cyanide heap leaching technology. Much of the test work was done at McClelland Laboratories, Inc., (MLI) and Colorado Minerals Research Institute (CMRI). A total of 32 column leach tests exist for Dolores; 22 different composites have been tested in three different metallurgical campaigns. Analysis of the column and bottle roll data by M3 and KCA predict the heap leach recoveries for Dolores ore crushed 80 percent minus 6.3mm as shown in Table 19-4.

Table 17-1
Predicted Metallurgical Recoveries
Ore Type Gold Recovery Silver Recovery
Oxide 79% 46%
Mixed 79% 52%
Sulfide, High Grade 65% 71%
Sulfide, Low Grade 56% 55%
Manganese Latite 79% 21%

 

17.1 Process Flow Sheet

Ore material is fed to the primary crusher at a nominal rate of 18,000 tonnes per day. Current average crusher availability is 17,000 to 17,500 tonnes per day. The crushing circuit is a three stage crush with recirculating load. Crushed material is fed via an overland conveyor system to the leach pads, where transport and stacking are done using grasshoppers and a radial stacking system. Lime is added to the crushed material during transport to the pads.

Crushed ore is leached on the leach pads using a combination of drip and sprayer systems. More sprayers are used during the rainy season (July – October) to increase evaporation in the circuit, while drip systems are used exclusively during dryer times of the year. Pregnant solution is collected in a PLS pond, oxygenated, clarified, and run through a Merrill-Crowe circuit.

In the Merrill-Crowe circuit, zinc dust is added to precipitate the gold and silver from solution. The solution is pumped to filter presses, where the resulting cake (excess zinc, gold and silver) are dried. Filter cake is further dried in a mercury retort to remove excess moisture and collect any mercury in the system. The dried cake is mixed with fluxes and then melted in a furnace. Slag material is decanted off from the furnace, leaving gold and silver, which is poured into dore bars.

76  

 




Bars are cleaned, assayed, weighed, and shipped to Johnson Matthey for final recovery. Dore bars from Dolores contain between 1 and 5% gold with 95-99% silver, and generally less than 1% impurities.

17.2 Actual Recovery Results

Ongoing process test-work and column tests are part of metallurgical review for the project. Recovery performance on the leach pad has been shown to be sensitive to solution flow rates and to cyanide concentrations, particularly for silver. Silver has a leach curve extending 430 days, which increases its reliance on secondary leaching of solution through subsequent lifts.

Actual metal recoveries track consistently with projected recoveries for gold. Silver recoveries in the Phase 1 leach pad appear to have been limited by reduced solution flow rates and cyanide concentrations. The metal remains in inventory and the remaining recoverable silver is expected to be extractable when the pad resumes irrigation. Silver recoveries in the phase 2 pad are currently tracking at 92% of projections. The minor delay in silver recovery is inferred to be related to solution saturation levels rather than any reduction in overall silver recoverability.

 

Figure 17-1
Cumulative gold recovery, phase 2 leach pad, relative to modeled gold recovery.

77  

 




 

Figure 17-2
Cumulative silver recovery, Phase 2 leach pad, relative to modeled silver recovery.

The recoveries appear reasonable, based on the data provided to CAM by Minefinders.

78  

 




18.0 PROJECT INFRASTRUCTURE

The Dolores project is located at a remote site in the mountains of western Chihuahua. Primary access is via an improved, unpaved road to Yapachic, approximately 50km to the south, where it links up with the main Chihuahua-Hermosillo highway. There is an alternate unpaved road access to the east, approximately 90km to Madera.

Primary camp facilities consist of 24 units of 6 rooms each, along with 12 2 room units, for a total of 168 rooms. Camp capacity stands at around 400 persons. There are separate man camps maintained by certain contractors which house contract personnel. Site services include laundry service, food preparation & service, recreational space, etc.

There is a fully permitted airstrip at the project which is capable of handling single-engine aircraft.

Power generation on site is by generator power. There is currently no connection to grid power. Primary power generation is by 6 1.8MW generators (generally 4 online, 1 stand-by, and 1 service generator.) The camp facilities have separate power generation.

Process water is sourced by pumping from the Tutuaca River, by pit dewatering activities, and from a storm-water management dam located in the Chabacan arroyo.

Primary telecommunications services are provided by overland microwave link. There is a backup satellite service available for voice and data, and limited cellular service available for voice.

Mine infrastructure: Primary mine-site infrastructure includes the primary crushing circuit, phase 1 and phase 2 leach pads (phase 3 leach pad to be constructed), waste stockpiles, pregnant solution and overflow ponds, Merrill-Crowe plant, Power generation plant, truck shops, warehouses, and offices.

79  

 




19.0 MARKET STUDIES AND CONTRACTS

Gold and silver are both precious metals which are sold at international market prices. Precious metals markets are elastic and are expected to absorb this production easily.

The Dolores mine and processing facilities produce a high grade silver-gold doré in bars assaying approximately 97.5% silver and 2% gold which is readily marketable. A refining agreement is in place with Johnson Matthey of Salt Lake City, Utah, which governs transport and final refining of dore from Dolores. Rates and charges for this contract are in line with industry norms.

There are a number of other vendor contracts in place at Dolores, including down-the-hole blasting support, access road maintenance, RC drilling services, camp services, food services, site transportation, etc. Rates, charges, and terms for all such contracts are in line with industry norms.

Production from the Dolores mine is unhedged.

80  

 




20.0 ENVIRONMENT STUDIES, PERMITTING AND SOCIAL OR COMMUNITY IMPACT

As discussed in Section 4, the relocation of Dolores villagers was completed in 2009.

20.1 Permitting

All environmental permits required for operations at Dolores are on file with the respective government agencies. Principal permits are as follows.

A. Environmental Impact Authorization (Autorización de Impacto Ambiental)

Permit number: S.G.E.A-DGIRA.-DDT.-0792.06
Issuance date: April 25, 2006
Issuing authority: Ministry of Environmental and Natural Resources (SEMARNAT)
Term: 18 years

B. Authorization for the Change of Land Use

Permit number: SG.CU.08-2006/034
Issuance date: April 17, 2006
Issuing authority: Ministry of Environmental and Natural Resources (SEMARNAT)
Term: Expires on December 31, 2018

C. Environmental Operating License (Licencia Ambiental de Funcionamiento)

Permit number: LAU-CHIH. 04/2009
Issuance date: October 12, 2009
Issuing authority: Ministry of Environmental and Natural Resources (SEMARNAT)
Term: Permanent license

D. Clean Industry Certification

Issuing authority: The Federal Office of Environmental Protection (PROFEPA)
Term: Expires on March, 2012

E. Authorization for the Extraction of Superficial Water

Permit number: BOO.00.RO3.04.2.-7032
Issuance date: December 12, 2007
Issuing authority: CONAGUA (National Water Commission)
Term: 20 years

The Federal Office of Environmental Protection (Procuraduría Federal de Protección al Ambiente) (“PROFEPA”) performed an environmental audit through which CMD obtained the certification as

81  

 




“Clean Industry” described in Section I.1.D above. Such certification endorses the legality and compliance of CMD’s operations with the applicable Mexican environmental laws and the Permits.

20.2 Reclamation

An estimate of reclamation costs is prepared on a quarterly basis. Minefinders is using an accretion of assets in the cash flow model to account for the cost of the eventual remediation.

Minefinders has submitted operating plans for and obtained permits to conduct all aspects of Dolores operations.

Minefinders plan to perform concurrent reclamation of the mine waste dumps by re-contouring the waste dump slopes to a 2.5:1 slope during normal operations. The cost for the mine dump reclamation is included in the mine operating costs. Dozers will perform the bulk of the work. When the last leach period is completed, the ore heap will be allowed to drain down and solution quality will be monitored for compliance.

Appropriate environmental action will be taken as solution quality during drain down is obtained. This could include rinsing, cyanide destruction, or microbial treatment to bring solutions into compliance. The pad will be “armored” with 0.5 meters of run of mine waste. Re-vegetation of the waste dumps and leach pad will be by applying seed bed preparation, seed, and fertilizer. $30,000 per month ($360k annually) has been included over the life of the mine for the revegetation. Other provisions for closure include pit area fencing, building and foundation demolition, an evapotranspiration cell (ET Cell), and ongoing monitoring. The pit will be allowed to flood.

There does not appear to be an acid rock drainage (ARD) issue with respect to waste rock disposal. The material was tested according to accepted industry practice and Mexican NOMs. Acid Base Accounting (ABA) showed that about 28% of the waste rock has a potential to generate acid with an ABA ratio of <1.2. A 20-week duration dynamic (humidity cell) test showed that only one of four composites made from ABA ratio <1.2 material generated a mildly acidic (pH = 5.8) leachate. That single composite represents only about 5.2 percent of waste material. The other three components made from ABA ratio <1.2 generated neutral or near neutral leachate.

Minefinders estimates the cost of reclamation to be $9.5 million and these costs are nearly offset by the estimated $7.5 m salvage value of plant equipment at the end of the mine life.

82  

 




21.0 CAPITAL AND OPERATING COSTS

Sustaining Capital costs for the project are primarily related to leach pad construction costs, and to addition and replacement of mining equipment during the remaining life of the mine. Replacement equipment prices reflect costs as of February 2011, which are substantially escalated relative to the 2008 feasibility work.

Outstanding sustaining capital costs are listed in Table 21-1.

These sustaining capital costs do not include pre-stripping amounts which may be capitalized under international financial reporting standards.

Table 21-1
Oustanding Sustaining Capital Costs
Sustaining Capital Amount (US$) Timeframe
Lab Equipment 100,000 Life of Mine
LIMS Installation 90,000 2011
Communications Equipment 100,000 2011-2012
Haul Trucks 46,200,000 Life of Mine
Mine Mobile Equipment 24,300,000 Life of Mine
Blasthole Drills 900,000 2011
RC Drill Equipment 1,000,000 2011
Leach Pad Construction (Phase 3 Pad) 30,000,000 2011-2013
Light Vehicles 1,000,000 Life of Mine
Maintenance Tools 750,000 2011
Chabacan Dam 4,721,000 2011
Camp Facilities 860,000 2011 – 2015
Process Plant improvements 1,600,000 2011

 

83  

 




22.0 ECONOMIC ANALYSIS

As noted in Section 3, an economic analysis has not been included based on a legal opinion provided to CAM.

22.1 Exploration Potential

Additional exploration potential exists at Dolores in three principal areas.

Deep mineralization below the current pit limit, where Drill intercepts below the current pit limits have returned results which with further testing, may prove to be of high enough grade to support underground development. The deep mineralization is contained within a multilithic breccia, and consists of galena, sphalerite silver sulfides, electrum, and native gold occurring in the matrix and open spaces of the breccia. Further drill testing will attempt to determine continuity of grade and mineralization. Due to the depth, high grades, and narrow expression of these zones, it may be preferable to continue exploration on these areas through underground exploration.

Drilling in the East Dike satellite deposit have returned ore-grade results over minable widths. The exposed mineralization consists of structurally controlled zones of stockwork veining hosted by intrusive latite, with drill results indicating wider zones at near surface depths. A portion of the east dike is contained within the current pit design. Further drilling may allow for extension of this satellite deposit and its addition into future mine plans. The East Dike satellite deposit is located within 200 meters of the east flank of the current pit. There is an additional deposit with ore-grade drill results contiguous with the East Dike zone to the north. This area is referred to as the North Dome. Additional drilling is planned during 2011 to expand resource in this area. This may allow for further extension of the East Dike satellite pit to the north.

Drilling indicates continuation of the main Dolores mineralized zones to the south beyond the area of the current pit. Mineralization there consists of stockworks and vein-style high-grade mineralization. Limited drilling in this area has indicated expanded zones at depth, with further testing required to determining overall grades and tons. Further expansion of mineralization in this area would potentially allow for extension of the pit to the south.

84  

 




23.0 ADJACENT PROPERTIES

The entire Dolores project area discussed in this report is controlled by Minefinders. All samples used in estimation of the mineral resources and reserves in this report were collected from within the Dolores property boundary. All the mineral resources and reserves disclosed in this report lie within the three Dolores concessions.

85  

 




24.0 OTHER RELEVANT DATA AND INFORMATION

The authors are not aware of any additional information, the exclusion of which would tend to make this report misleading.

86  

 




25.0 INTERPRETATION AND CONCLUSIONS

Based on a review of the database, resource estimation methodology and reconciliation provided by Minefinders and additional independent calculations by CAM, CAM believes that the resource model for Dolores has been prepared according to accepted engineering practice and is suitable for public disclosure and use in reserve calculations and financial planning.

There is only a limited amount of production data under the new mining method and the resource estimate may need to be revised as more production experience with the new mining method is gained. CAM has made recommendations to Minefinders for improvement in resource estimation methodology.

87  

 




26.0 RECOMMENDATIONS

The Dolores property is in commercial production, and CAM did not provide a Work Program for further exploration. The following recommendations relate to database and other routine ongoing activities.

  1.

Correct errors identified by CAM in the Exploration and RC databases. Although it appears that these errors do not affect the global resource estimate they may be significant locally and should be corrected.

2.

The current procedure of reviewing QA/QC data for RC duplicates as a simple percentage difference is not best engineering practices as it generally tends to flag low grade samples and ignore errors at the higher grade. A simple scatter plot of original versus replicate value on both in untransformed and log-log basis should be adequate for this check.

3.

A formal report on QA/QC on the drilling since the last (2008) report needs to be prepared.

4.

The CAM sector search check model gave fewer ounces than either of the two Minefinders models which did not use a sector search. Using a sector search at least as a sensitivity run is a good check to avoid central higher grade composites possibly over driving the pit.

5.

There are some cases (e.g. the x-coordinate for composites) where the maximum and minimum specified for the fields in MineSight are less or greater than the actual maximum and minimum of the data. This appears to have had no affect on the resource estimate but should be corrected.

6.

The multi-run feature in MineSight is a powerful and useful tool for developing complicated models. However, some multiple runs overwrite the report files from prior runs render E. in impossible to properly audit the preparation the model. For this reason CAM recommends that each multi-run report file have a unique name.

7.

The probabilistic model lacks the tight constraints of the interpreted structures used in the constrained inverse distance squared model and has significantly wider across strike search than the CAM checked models. This means that there may be risk of a few high-grade composites at the bottom of the hole, which locally result in the prediction of too much metal. It is a relatively simple matter to calculate the amount of metal associated with each hole and composite particularly after recommendation six is implemented.

8.

As of the close-off date, there was insufficient data to completely validate the model against the new model. After 1 year of production with the new ore control method the reconciliation should be reviewed and the exploration model revised as necessary.


88  

 




27.0 REFERENCES

Albinson, T., Norman, D.I., Cole, D., and Chomiak, B., 2001, Controls on formation of low-sulfidation epithermal deposits in Mexico: Constraints from fluid inclusion and stable isotope data: Soc. Econ. Geol. Special Pub. 8. p. 1-32.

Ball, Shaun, Page, Tench, and Bailey, M.H., November 2006, Technical Report on the Mineral Reserve Update for the Dolores Property, Chihuahua, Mexico: internally prepared NI 43-101 format report, Minefinders Corporation Ltd.

Chlumsky, Armbrust and Meyer, LLC (CAM), July 2007, NI 43-101 Technical Report – Dolores Gold-Silver Project, prepared for Minefinders Corporation Ltd. by R. L. Nielsen and R.L. Sandefur, 80 pages in .pdf format .

Corbett, G.J. and Leach, T.M., 1998, Southwest Pacific Rim gold-copper systems: Structure, Alteration, and Mineralization: Soc. Econ. Geol. Special Pub. 6, 240p.

Gustavson Associates, 2008, NI 43-101 Technical Report on the Mineable Reserve for the Dolores Gold-Silver Project, Chihuahua State, Mexico: consulting report dated March 25, 2008 by Gustavson Associates, prepared for Minefinders Corporation Ltd., 94 pages in .pdf format.

KCA, 2006, Technical Report for the Dolores Heap Leach Project in Mexico: consulting report dated April 11, 2006 by Kappes Cassiday Associates, prepared for Corporation Ltd., 94 pages.

KCA, 2006, Metallurgical consulting report dated April 11, 2006 by Kappes Cassiday Associates, prepared for Minefinders Corporation Ltd.

M3 Engineering & Technology Corp., June 2005, Dolores Feasibility Study, Vol. 1, NI 43-101 Technical Report for Dolores Project, Chihuahua: unpublished report for Minefinders Corporation Ltd.

Minefinders Corporation Ltd., 2006, Technical Report on the Mineral Reserve Update for the Dolores Property, Chihuahua, Mexico: internally prepared NI 43-101-format report, revised November 7, 2006, by Minefinders Corporation Ltd. 68 pages with appendices.

MRDI Canada, October 1998, Dolores Scoping Study: unpublished report for Minefinders Corporation Ltd.,

Overbay, W.J., Page, T.C., Krasowski, D.J., Bailey, M.H. and Matthews, T.C., 2001, Geology, Structural Setting, and Mineralization of the Dolores District, Chihuahua, Mexico: in New Mines and Discoveries in Mexico and Central America, Tawn Albinson and C.E. Nelson, edts., Society of Economic Geologists Special Publication No. 8, 362pp.

PAH, 2002, Audit of Resources at the Dolores Gold-Silver Project, Chihuahua, Mexico: unpublished consulting report dated December 2002, prepared for Minefinders Corporation Ltd. by G.A. Armbrust and R.L. Sandefur of Pincock, Allen & Holt.

89  

 




Roscoe Postle, 2004, Technical Report on the Mineral Resources Estimate for the Dolores Property, Mexico; unpublished NI 43-101 report dated November 2004, prepared by Roscoe Postle Associates, Inc. for Minefinders Corporation Ltd. 80 pages.

90  

 




28.0 DATE AND SIGNATURE PAGE

I, Robert L. Sandefur

1139 S Monaco Pkwy
Denver, CO 80224

hereby attest that:

a)

I am a Consulting Geostatistician, affiliated with Chlumsky, Armbrust and Meyer LLC at 12600 W. Colfax Avenue, Suite A-250, Lakewood, Colorado 80215, USA.

b)

I am a Certified Professional Engineer (Number 11370) in the state of Colorado, USA, and a member of the Society for Mining, Metallurgy, and Exploration (SME).

c)

I graduated from the Colorado School of Mines with a Professional (BS) degree in engineering physics (geophysics minor) in 1966 and subsequently obtained a Masters of Science degree in physics from the Colorado School of Mines in 1973.

d)

I have practiced my profession as a geostatistical resource analyst continuously since 1969. From 1969 to present, I have worked on mining projects in over 20 countries, have statistically analyzed more than 400 mineral deposits, and have personally visited more than 50 operating metal mines.

e)

I have read the definition of “qualified person” set out in National Instrument 43-101 (“NI 43-101”) and certify that by reason of my education, affiliation with a professional association (as defined in NI 43-101) and past relevant work experience, I fulfill the requirements to be a “qualified person” for the purposes of NI 43-101.

f)

I am responsible for the preparation of sections 14 and 17 and the relevant portions of sections 1, 20, 21, and 23, of the report entitled "NI 43-101 Technical Report, Dolores Gold-Silver Project, Chihuahua, Mexico, dated 30 September 2011, and for which the effective date is 31 December 2010.

g)

I visited the subject property on 23-24 September 2010.

h)

As defined in Section 1.5 of National Instrument 43-101, I am independent of the issuer, Minefinders Corporation, Ltd.

i)

I am not aware of any material fact or change with respect to the subjects of this report which is not reflected in this report, the exclusion of which would make this report misleading.

j)

I have read National Instrument 43-101 and Form 43-101F1, and the report has been prepared in compliance with that Instrument and Form.

k)

I hereby notify the British Columbia Securities Commission of my consent to the filing of this Technical Report with stock exchanges and other regulatory authorities in Canada, and any publication by them, including electronic publication in the public company files on their website accessible by the public, of the Technical Report.


91  

 




Dated this 30th day of September, 2011

Signed /s/ R. L. Sandefur
Robert L. Sandefur, P.E.

92