Pampa el Toro Mineral Resource Technical Report

Report Prepared for

Cardero Resource Corp

Report No360457/1

September 10, 2009

Pampa el Toro Mineral Resource Technical Report

Cardero Resource Corp

Suite 1920 - 1188 West Georgia Street
Vancouver, BC
Canada, V6E 4A2

SRK Project Number 360457

265 Oxford Road

Illovo

2196

South Africa

P O Box 55291

Northlands

2116

South Africa

Tel: (011) 441-1111

Fax: (011) 880-8086

Mark Wanless mwanless@srk.co.za

September 10, 2009

Compiled by:

Reviewed by:

Mark Wanless

Sean Meadon

Victor Simposya

Project Consultants

Partner

Table of Contents

Location

Ownership

Geology and Mineralisation

Exploration Concept

Status of Exploration

Mineral Resource Estimation

Qualified Person’s Conclusions and Recommendations

2.1

Qualifications of SRK

2.2

Terms of Reference

2.3

Purpose of Independent Technical Report

2.4

Site Visit

2.5

Sources of Information

4.1

Location

4.2

Land tenure

4.3

Environmental and Socio-Economic Issues

5.1

Accessibility

5.2

Climate

5.3

Local Resources and Infrastructure

5.4

Physiography

10.1

Surface Sand Sampling

10.1.1

Reconnaissance Surface Sand Sampling Phase

10.1.2

Infill Surface Sand Sampling Phases I and II: Pampa El Toro

10.2

Trenching

11.1

Percussion Drilling

11.1.1

Percussion Drilling Programme:  Phase I

11.1.2

Percussion Drilling Programme:  Phase II

12.1

Reconnaissance and infill surface sand sampling

12.1.1

Field collection methods

12.1.2

Sample documentation and security

12.1.3

Quality control procedures in the field

12.2

Percussion Drilling

12.2.1

Field collection methods

12.2.2

Sample documentation and security

12.2.3

Quality control procedures in the field

12.3

Trenching

12.3.1

Field collection methods

12.3.2

Sample documentation and security

12.3.3

Quality control procedures in the field

12.4

Bulk Sample Collection

12.4.1

Field Collection Methods

12.4.2

Sample documentation and security

12.4.3

Quality control procedures in the field

13.1

Magnetic Separation and processing

13.1.1

Magnetic separation and processing of reconnaissance surface sand samples

13.1.2

Magnetic separation and processing of drillhole and infill surface sand samples

13.1.3

Quality control and security

13.2

Sample analysis and assay procedures

13.2.1

Sample Preparation

13.2.2

Analysis

14.1

Cardero Verification

14.1.1

Quality Control Procedures

14.1.2

Internal and External Check Assays

14.2

SRK Verification

16.1

Mineral Processing and Metallurgical Testing 2005 - 2007

16.15.1

SGS Lakefield (2005)

16.15.2

Midrex Test (2005)

16.15.3

Solumet Upgrading Test (2006)

16.15.4

Eriez Magnetic Separation Test (2006)

16.15.5

Midrex Test (2006)

16.15.6

Bateman Engineering Test (2006)

16.15.7

Bateman Engineering Test (2007)

16.16

Mineral Processing and Metallurgical Testing 2008

16.16.1

Mineral Processing

16.21.1

Metallurgical Testing

17.1

Data Statistics

17.2

Semi Variogram analysis

17.3

Density Determination

17.4

Wireframe modeling

17.5

Resource estimation

17.6

Classification and Mineral Resource Reporting

Mark D. Wanless

L. Holland

List of Tables

Table ES1.1:

Pampa el Toro Mineral Resources as at Mineral Resources as at 21 July 2009.¹

Table 4.1:

List of existing concessions as relating to the Pampa el Toro Project

Table 14.1:

Certified values of selected variables for OREAS 42P and statistics of assay results

Table 14.2:

Certified values of selected variables for GPAB-6 and statistics of assay results

Table 14.3:

Descriptive statistics of the duplicate datasets for a range of analysed variables.

Table 16.1

Summary of source and results of tests conducted on five bulk samples from Pampa el Toro bulk samples from Carbonera – 2005-2007

Table 16.2

Summary table of amount of sand processed, magnetic concentrate produced, and two weight recoveries and concentrate grades from each of the three pilot plant tests. (std = standard deviation)

Table 16.3

NRRI in-house laboratory results from the screening test on Pampa el Toro magnetic concentrate.

Table 17.1:

Univariate statistics of selected variables from the 5m composites.

Table 17.2:

Correlation Matrix of selected variables from the 5m composites.

Table 17.3:

Modeled semi-variogram parameters for all variables estimated

Table 17.4:

Mineral Resources for the Pampa el Toro project as at 21 July 2009

List of Figures

Figure 4.1:

Location map of the Pampa el Toro project

Figure 4.2:

Map of Pampa el Toro properties and dune field boundary.

Figure 10.1:

Pampa el Toro reconnaissance surface sand sampling, distribution and results.

Figure 10.2:

Pampa el Toro infill surface sand sampling, phases I and II, distribution and results.

Figure 11.1:

Map of percussion drill hole collars.

Figure 11.2:

Map of percussion drillhole duplicate and trench locations.

Figure 12.1:

Schematic diagram showing field procedure for collecting reconnaissance sand samples at pampa sand sites.

Figure 12.2:

Schematic diagram showing field procedure for collecting reconnaissance sand samples at dune sand sites.

Figure 12.3

Schematic diagram showing procedure for collecting field duplicate samples during reconnaissance sand sampling at dune sand sites.

Figure 12.4

Schematic diagram showing procedure for collecting within-pit duplicate samples during reconnaissance sand sampling at dune sand sites.

Figure 12.5

Typical 20-sample collection “block” used in The Pampa El Toro reconnaissance surface sands sampling programme.

Figure 12.6

Typical quality control scheme used in the sampling of sands from a typical 100 metre-deep percussion borehole at Pampa El Toro.

Figure 14.1:

Time sequence control plot of analytical results of OREAS 42P and GPAB-6 for Fe₂O₃ %

Figure 14.2:

Time sequence control plot of analytical results of OREAS 42P and GPAB-6 for

SiO₂ %

Figure 14.3:

Time sequence control plot of analytical results of OREAS 42P and GPAB-6 for

TiO₂ %

Figure 14.4:

Time sequence control plot of analytical results of GPAB-6 for V ppm

Figure 14.5:

Scatter plots of Fe₂O₃ duplicate analyses from Splitter, Preparation and Pulp datasets

Figure 14.6:

Scatter Plot of SiO₂ original and duplicate pulp analyses

Figure 14.7:

Scatter Plot of Fe₂O₃ original and duplicate pulp analyses

Figure 14.8:

Scatter Plot of TiO₂ original and duplicate pulp analyses

Figure 14.9:

Scatter Plot of V original and duplicate pulp analyses

Figure 14.10:

HARD Plot of SiO₂ duplicate analyses from the pulp dataset

Figure 14.11:

HARD Plot of Fe₂O₃ duplicate analyses from the pulp dataset

Figure 14.12:

HARD Plot of TiO₂ duplicate analyses from the pulp dataset

Figure 14.13:

HARD Plot of V duplicate analyses from the pulp dataset

Figure 16.1:

Illustrated photograph of the Eriez magnetic separation pilot plant.

Figure 16.2

Flow sheet illustrating the magnetic separation process of the pilot plant.

Figure 17.1:

Histograms of grade variable for selected variables from the 5m composites.

Figure 17.2:

Scatter plots of Fe₂O₃, TiO₂, V and SiO₂ from the 5m composites indicating the strong correlations that exist in the data.

Figure 17.3:

Directional semi-variogram for Fe₂O₃ in the plane approximately parallel to the topography

Figure 17.4:

Plan view of drillhole collars displaying masked drillholes in red

Figure 17.5:

Omni-directional semi-variograms in the plane parallel to the topography surface, for Fe₃O₂, TiO₂, V and SiO₂

Figure 17.6:

Cross Validation outputs for Fe₂O₃ %, from the 5m composites.

Figure 17.7:

Cross Validation outputs for TiO₂ %, from the 5m composites.

Figure 17.8:

Cross Validation outputs for Vanadium ppm, from the 5m composites.

Figure 17.9:

Histogram of compacted (left) and un-compacted (right) Bulk Density determination results

Figure 17.10:

Kriging search neighbourhood analysis results

Figure 17.11:

Schematic illustration of the samples selected in a typical search neighborhood

Figure 17.12:

Classified Mineral Resources with drillhole collar locations

10 September 2009

360457

Pampa el Toro Mineral Resource Technical Report

1

Summary

Cardero is a leader in exploration and development of Iron-ore, Copper and Gold projects in the Americas. Their projects include a highly prospective Iron Oxide Copper Gold (IOCG) project within the Alisitos Arc in Baja California Norte, Mexico, various gold projects in Argentina and Iron Sands projects in Peru. The Pampa el Toro Project is an Iron Sands project situated in the desert coastal region of Southern Peru. During late 2005 Cardero initiated a feasibility study on this project by starting a sand sampling program headed by Cooke Geochemical Consulting, and a trenching and percussion drilling programme managed by their own geologists, This has been followed up by various outsourced metallurgical test programs, initially done by Midrex (2005) and most recently by Bateman Engineering (2007).

Location

The property is located near the city of Nazca in the desert coastal region of southern Peru approximately 45 km northeast of the port of San Juan, close to the large Marcona iron mine (approximate resource of 1.4Bt @ 54% iron), the Mina Justa copper deposit (inferred resource of 218Mt @ 0.8% copper), and to Cardero's Pampa de Pongo iron deposit

Ownership

There are 14 mining concessions within the Pampa El Toro Project area totalling 10 00 hectares. Cardero has direct ownership of the 9 concessions  and has signed an option agreement with Minas Ataspacas S.A. to acquire an initial 70% interest in the other 5 concessions.

Geology and Mineralisation

The Lower Paleozoic iron rich Marcona Formation is unconformably overlain by Triassic, Jurassic and Cretaceous volcanic and sedimentary sequences which, in turn, were intruded by stocks and dykes belonging to the Peruvian Coastal Batholith complex.  These strata were, in turn, deformed, uplifted and eroded, and formed a base on which Quaternary sediments were deposited.

It is speculated that the iron-bearing minerals contained in the iron sands either partially or wholly originated from the volcanic units overlying the Marcona formation, however additional as yet unidentified sources are possible. However, it is possible that erosion of these Quaternary sediments contributed significant quantities of magnetite material contained in the dune fields.

Mineralogically, the principal iron-bearing minerals in the dunes consist of magnetite, titanomagnetite and ilmenite, with minor to trace amounts of hematite, rutile, titanite, ulvospinel and goethite.  Silicate minerals, which make up the bulk of the sands, are primarily quartz, plagioclase, amphibole, chlorite, mica and pyroxene (SGS Lakefield Research, 2005a).  The majority of the ilmenite (80-85% by volume) is associated with magnetite, titanomagnetite and hematite, with the remaining 15-20% by volume present as free ilmenite grains.

The dune sands are considered to be the product of aeolian re-distribution of contemporary and ancient beach sands and Quaternary basin sediments derived from erosion of the bedrock, winnowed, and re-distributed, by prevailing winds (Gay, 2005) to produce potentially economic concentrations of magnetite in extensive, and partially active, dune fields.

Exploration Concept

The primary focus of the exploration work has been on the sampling, trenching and percussion drilling of the deposit, combined with significant amounts of metallurgical testwork. This has lead to the development of a robust geological and structural model on which the mineral resource estimation and calculation is based. Exploration is thus at an advanced stage. Detailed metallurgical testwork has also been completed.

Cardero has conducted the exploration program with a set of standard procedures which aim to monitor the quality of the sampling and assay results. The standard procedures include submitting both Certified Reference Materials to monitor the accuracy of the analyses, as well as the analysis of field and laboratory pulp duplicate samples to benchmark the sampling and sample splitting errors as well as the precision and repeatability of the analyses.  The quality control samples are checked before accepting the batch analytical results from the laboratory in order to control the quality of the data accepted in the exploration database. The quality control samples indicate that there was no bias introduced in the samples splitting process, as field duplicates and pulp duplicates from the laboratory show very similar characteristics. Analysis of the quality control sample results indicates that the analyses have been conducted to a high level of accuracy and precision, and are acceptable for use in Mineral Resource Estimation.

Status of Exploration

A pilot mineral processing plant has been erected on site in early 2008 and approximately 56.8 tons of magnetic concentrate was produced. In terms of metallurgical testwork The Phase I briquetting work and the Phase II bench-scale pig iron production program has been completed. The Phase III pilot-scale pig iron production program has been completed and the analytical results are currently being determined.

Mineral Resource Estimation

SRK have generated a Mineral Resource estimate based on the data supplied by Cardero from their exploration programs. The assay results of a selected set of elements from the 5m composite samples were analysed and found to represent a relatively well represented single population of all elements. There are strong correlations between the major economic elements (Fe₂O₃, TiO₂, and V) that indicate a common genetic and depositional history. This indicates that these elements are most likely contained within the same, or strongly associated minerals.

SRK generated and modeled experimental semi-variograms that indicate long range continuity in the major elements, but that also indicate a long range trend element within the deposit in a north-south direction. Cross validation tests conducted by SRK on the semi-variograms indicate that the modelled semi-variograms and the selected search neighbourhood parameters should provide robust estimates.

The surface topography contours were used to generate a wireframe representation of the topography. This surface was copied vertically downwards by 30m, and used to constrain the depth extent of the Mineral Resource estimate. Drillhole collars were also projected vertically onto the surface as the various generations of GPS surveys of the collars did not accurately match the surveyed contours.

SRK created a block model with a block size equivalent to the nominal drillhole spacing of 250m X and Y directions, and 5m vertically. The block model was rotated to approximately match the dip of the topographic surface, which coincided with the plane of best continuity. In order to better model the volume, and the topographic variations, the parent blocks were sub-divided into smaller blocks, with a minimum size in the XY plane of 6.25m by 6.25m. The blocks were created to exactly match the intersection of the wireframe with the block center. Only parent blocks were estimated (i.e. each 250m by 250m by 5m collection of sub-blocks will have the same grade estimate)

The vertical continuity of mineralisation is significantly shorter than the lateral continuity, as would be expected from a deposit emplaced and reworked in thin sheets over time. The search ellipsoid employed by SRK takes this into account using anisotropic search scaling, with the result that samples from adjacent boreholes are given a higher weighting than samples from the same borehole  that occur above or below the block being estimated. A minimum of four samples was required to estimate a block, and a maximum of 12 was used, to preserve some local variability in the grades. The maximum number of samples used in the estimation was selected after an optimisation exercise indicated that this was the best balance between generating high quality estimates, and over-smoothing of the estimation.

The Mineral Resources were classified on the basis of the confidence in the geological variation, the quality of the sampling and analytical results, drillhole spacing, and indicators of the quality of the estimation. The central portion of the deposit, which is drilled on approximately 250m centers, which has high quality estimates was classified as in Indicated Mineral Resource. The portion of the deposit surrounding this, drilled on approximately 500m centers, and extending approximately 500m beyond the area drilled on 500 centers is classified as an Inferred Mineral Resource. Portions of the deposit that were estimated further than 500m from the 500m spaced drilling were also excluded from the mineral Resources until further confirmatory work is completed to confirm the grades of the material. The Mineral Resources are reported in Table ES1.1.

Table ES1.1: Pampa el Toro Mineral Resources as at Mineral Resources as at 21 July 2009.¹

Category	Volume (m3 '000)	Tons (kT)	Grade (Fe2O3 %)	Grade (TiO2 %)	Grade (V ppm)
Indicated	133,608	241,831	6.67	0.72	172
Inferred	348,190	630,224	6.47	0.70	166

1 No grade cut-off has been applied to the material in the Mineral Resource, as the grades are relatively homogenous, would likely be able to be blended, and are projected to be economically viable, based on the assumptions made by Cardero. A Mineral Resource is not a Mineral Reserve and does not have demonstrated economical viability.

Qualified Person’s Conclusions and Recommendations

In the Exploration conducted to date at the Iron Sands Project, the surface sampling and subsequent percussion drillhole testing indicates that the overall magnetite content is relatively homogeneous, particularly in the uppermost 30 meters.  Sample density achieved in the investigations is considered to be more than adequate to support these assertions.

The underestimation of magnetic mineral content and Fe grades caused by milling and grinding of sand is not expected to affect the in situ Fe grade which the current Mineral Resource estimate is based.  

As an additional upgrading step to dry magnetic separation, screening of the concentrate is both simple and economic. Further testing in this regard is strongly recommended

A robust Mineral Resource estimate has been generated based on a sound geological model and exploration database. The Mineral Resource is classified in terms of the CIM definitions on Standards on Mineral Resources and Reserves – Definitions and Guidelines. An Indicated Mineral Resource of 133,608 kT with a Fe₂O₃ grade of 6.67% has been defined, along with an Inferred Resource of 348,190 kT with a Fe₂O₃ grade of 6.47%.

SRK recommend that the metallurgical test work continue to improve the confidence in the ability to upgrade the raw sand material into a saleable product, and that Cardero proceed with the preliminary economic assessment of the project

2

Introduction

SRK Consulting South Africa (SRK) have prepared an independent Mineral resource estimate of the Pampa el Toro Iron sands project (the Project), based on information supplied to SRK by Cardero, as well as knowledge gained about the property during two site visits to the project by representatives from SRK. The Mineral Resource estimate was generated by Mr. Mark Wanless a Professional Natural Scientist (Pr.Sci.Nat) registered with the South African Council for Natural Scientific Professionals (SACNASP). SACNASP is a statutory body which is recognized as a Recognized Overseas Professional Organization (ROPO) by the Canadian Institute of Mining, Metallurgy and Petroleum. Mr. Wanless served as the Qualified Persons responsible for the preparation of the Independent Technical Report (ITR) as defined in National Instrument 43-101, Standards of Disclosure for Mineral Projects, and in compliance with Form 43-101F1 (the Technical Report).  

2.1

Qualifications of SRK

SRK is part of an international group (the SRK Group), which comprises more than 800 staff, offering expertise in a wide range of resource engineering disciplines.  The SRK Group’s independence is ensured by the fact that it holds no equity in any project, contractor or supplier.  This permits SRK to provide its clients with conflict free and objective recommendations on crucial judgement issues. 

The SRK Group has a demonstrated track record in undertaking exploration programmes, independent assessments of resources and reserves, technical due-diligence audits, competent person’s reports, project evaluations and independent feasibility evaluations to bankable standards on behalf of exploration and mining companies and financial institutions world wide. The SRK Group has worked on a large number of major international mining operations and projects for all the major stock exchanges and has specific experience in commissions of this nature.

This ITR has been prepared based on a technical review of the exploration processes, and an independent Mineral Resource estimation by SRK.  

Neither SRK nor any of its employees and associates employed in the preparation of this report has any significant beneficial interest in Cardero or in the assets of Cardero.  SRK will be paid a fee for this work in accordance with normal professional consulting practice.  The payment of this fee is not contingent upon the conclusions or opinions expressed in this report.

The individuals who have provided input to this ITR, who are listed below, have extensive experience in the mining industry and is a member in good standing of appropriate professional institutions.

• Mark Wanless, PrSciNat,  BSc (Hons)  -   Geology, Mineral Resources
• L. Holland, B.Sc., C.Eng., FIMMM - Mineral Processing Engineer

The Qualified Person with overall responsibility for the reporting of this ITR is Mark Wanless, PrSciNat, BSc (Hons), who is an employee of SRK. Mr Wanless is a mining geologist with 13 years experience in the mining industry and has been responsible for the reporting of Mineral Resources on various properties in Southern Africa and internationally during the past five years.

L. Holland, B.Sc., C. Eng., FIMM, is the Consultant Minerals Processing Engineer for Holland and Holland Consultants. He is the qualified person responsible for the mineral processing and metallurgical sections of this report (section 16). He is a Mineral Processing Engineer with 41 years experience in the mining and mineral processing industry, is a Qualified Person (QP) with special expertise in Metallurgy/Mineral Processing and is a fellow in good standing with the Institute of Materials, Minerals, and Mining (FIMM), UK.). He has visited the Pampa el Toro property for six days in May 2008 and four days in June 2008.

2.2

Terms of Reference

SRK was required to prepare an ITR for the Project according to the requirements of National Instrument 43-101 (NI 43-101) in support of the disclosure of a Mineral Resource estimate for the Project. The work involved auditing the exploration practices used to generate the source information, and estimating a Mineral Resource for the project.

2.3

Purpose of Independent Technical Report

This ITR was prepared to support the disclosure made by Cardero on 21 July 2009 of a Mineral Resource estimate for the Project.  It has been prepared in accordance with the requirements of NI43-101 and the NI43-101F1, and conforms with generally accepted Canadian Institute of Mining (“CIM”) “Exploration Best Practices” and “Estimation of Mineral Resources and Mineral Reserves Best Practices” Guidelines.

2.4

Site Visit

In compliance with NI 43-101 guidelines, Mr Mark Wanless, a Qualified Person, visited the Project in February 2008. During the site visit, SRK viewed a selection of drilling locations and the extent of the project area. SRK reviewed the sampling procedure as well as the processes used by Cardero to generate magnetic concentrates from the raw samples. SRK recommended that analyses should be done on the feed samples. The analyses were then used in the generation of a Mineral Resource estiamate. Cardero subsequently created a number of sub-samples from the reference samples stored at the drilling locations for the purposes of generating representative 5m composites for the top 30m of all drillholes.

In addition, the project has previously been visited in June 2006 by Mr Andre van der Merwe, employed by SRK at the time. Mr van der Merwe reviewed the drilling and sampling processes used by Cardero during the exploration which was ongoing at that time and made recommendations on improvements to the drilling and sampling processes that were implemented by Cardero.

2.5

Sources of Information

In preparing this report, the author relied on the technical report (2009) written by Dr. Cruise and Mr. Hoffman of Cardero, who in turn relied on various geological reports and maps, metallurgical and analytical reports and technical papers listed in the References section at the end of this report.  Additional information and data for review by Dr. Cruise and Mr. Hoffman in connection with the preparation of their Technical Report were obtained from the Instituto Nacional de Concesiones y Catastro Minero (INACC) in Lima and through consultations with Cardero staff, including Dr. Jayson Ripke, Vice President Technical, Cardero Iron Ore Company Ltd., and Tansy O’Connor-Parsons, Senior Geochemist, Cardero Resource Corp

3

Reliance on Other Experts

In preparing this report, SRK relied on geological reports and maps, metallurgical and analytical reports and technical papers listed in the References section at the end of this report. SRK Have relied upon Cardero for sections 3 to 8 and 15 of this report, and have based sections 9 to 12 on information supplied by Cardeo, and observations during site visits by SRK Additional information and data in connection with the preparation of this report were obtained from the Instituto Nacional de Concesiones y Catastro Minero (INACC) in Lima and through consultations with Cardero staff, including Dr. Jayson Ripke, Vice President Technical, Cardero Iron Ore Company Ltd., and Ms. Tansy O’Connor-Parsons, Senior Geochemist, Cardero Resource Corp.

SRK have not independently verified the legal status or ownership of any of the properties covered by this ITR, or of any related option agreements, nor have they verified the legal status of any joint venture arrangements covering adjacent third-party-owned properties.

4

Property Description and Location

4.1

Location

The Pampa el Toro Project (Figure 3.1) is located approximately 450 km southeast of Lima, and is 80 km southeast of Nazca in Peru.  The coastal village of Lomas, from which all Project–related field work was conducted, and which hosts the laboratory and storage facilities for the Project, lies 528 km by road from Lima and 7 km from the Panamericana Sur Highway. The magnetic separation pilot plant is located approximately 5 km due west from the center of the Pampa el Toro property, and next to the town of Acari. The Pampa El Toro Dune Field, lies within the confines of the Instituto Geografico Nacional (IGN) 1:100,000-scale topographic sheet 31-n (Acari).

Figure 4.1: Location map of the Pampa el Toro project

4.2

Land tenure

The 14 mining concessions which now constitute the Pampa el Toro property total 10,300 ha. The distribution of individual concessions and corresponding reference number is shown in Figure 3.2, and a complete list is provided in Table 3.1. The status of each of the concessions in Table 3.1 has been checked by Cardero against the INACC database (effective date June 30, 2009) and found to be correct.

Cardero has assumed, from a private Peruvian company, all rights and obligations under an agreement, dated 16 December, 2005, between the private Peruvian company and Minera Ataspacas S.A. (a Peruvian private company) whereby the private Peruvian company has the option to acquire from Minas Ataspacas S.A. an initial 70% interest in five mineral sands concessions (totalling 3,600 ha) situated adjacent to Cardero’s existing iron sands claims.  In order to exercise the option, Cardero is required to pay a total of US$ 2,500,000 over five years to 15 December, 2010, and incur exploration expenditures of not less than US$ 250,000 over the same period.  Upon Cardero having the acquired the 70% interest, a joint venture company will be formed with Minas Ataspacas S.A., and each party will thereafter be required to contribute its share of ongoing expenditure or be diluted.  If either party is diluted to less than 10%, such interest will be converted to a 2% NSR royalty.  If Minera Ataspacas S.A. is reduced to the 2% NSR, Cardero may purchase half the NSR (1%) for US$ 2,000,000 within 24 months of the exercise of the option, and the remaining half (1%) for US$ 8,000,000 within 36 months of the exercise of the option.

SRK has not independently reviewed the mineral titles to assess the validity of the stated ownership, and relies on documentation provided by Cardero.

Table 4.1: List of existing concessions as relating to the Pampa el Toro Project

* Reference N^o’s refer to locations shown in Error! Reference source not found.

4.3

Environmental and Socio-Economic Issues

The immediate property area is uninhabited, and the nearest settlement, Acari, subsists predominantly on agriculture. If water is determined to be necessary in the process of extracting the magnetite sands, the local population will need to be consulted.

Figure 4.2: Map of Pampa el Toro properties and dune field boundary.

5

Accessibility, Climate, Local Resources, Infrastructure and Physiography

5.1

Accessibility

Access to the Property is via the Panamericana Sur Highway from Lima (528 km to the coastal village of Lomas, approximately six hours by car), and then approximately 60 km by local dirt roads (around 1 hour on average; Figure 3.1) to the Pampa El Toro Dune Field.  It is accessed from the Acari Junction on the Panamerican Sur Highway and then via local dirt road to the village of Acari.  An access road was built from Acari to the upper levels of the dune field prior to percussion drilling, although access was restricted to dune-buggy or light pick-up trucks due to the steepness of the sandy slopes.  A heavy duty road capable of taking drilling equipment was constructed from the Panamericana Sur Highway just south of the village of Chavina in November 2005 (Figure 3.2).

5.2

Climate

The Project area lies within the desert coastal tract of southern Peru, a northward continuation of the Atacama Desert of northern Chile.  It is extremely arid, with less than 1cm annual rainfall and is virtually devoid of vegetation, except in irrigated river valleys.  During the summer months, the climate is extremely hot and dry, with only gentle breezes along the coast and high temperatures over the dune fields and valleys, with little movement of sand.

During the winter months, the coastal areas below 1,000 m.a.s.l. are cool, and covered by sea mists which contribute to a humidity of around 100% and light drizzle which gives rise to sparse vegetation (“lomas”) on which domestic animals are able to graze.  Strong onshore winde (southeast, verging-to-south, and-to-southwest) winds are generated at this time by solar heating of the hinterland pulling cooler air inland from the coastal belt.  These winds carry the mist and a considerable burden of aeolian sand with them, the latter being deposited against rising ground as the strength of the winds diminishes.

5.3

Local Resources and Infrastructure

The local economy is very poor, and dominated by agriculture and animal husbandry in the river valleys where irrigation is possible.  The dune fields constitute hostile terrain and are generally unvisited.

The deep-water port of San Juan (exit port for the Marcona Iron Mine) is situated around 40 km from the Panamericana Sur Highway and a further 100 km (maximum) from any of the dune fields under investigation.

5.4

Physiography

The large dune fields of southern Peru were described in detail by Gay (2005), who referred to the Pampa El Toro Dune Field (Acari Sand Mass) as resulting from the action of strong, persistent, onshore winds (southeast-to-south-to-southwest) acting on large quantities of coastal beach sands derived from rapid uplift of the Andes (and coastal belt) under the influence of an extremely arid climate. He concluded that the dune fields accumulated at elevations of between 500m and 2000m in the Andean foothills where weakening of the carrying winds against rising topography led to deposition of the sands.

The Pampa El Toro Dune Field comprises a mainly flat-lying to gently-rolling pampa or “erg” (sand sea) in its southern and eastern part, and active dune fields in the north and west.  

The Dune Field is oriented roughly north-south around the western and southern flanks of Cerro de Arenas.  Surface area amounts to > 120 km², of which around two-thirds in the south and south-eastern part of the field consists of pampa sands (a flat-to-gently-rolling sand apron surface similar to that of an “erg” or “sand sea”).  The remaining third consists of active dunes of small and similar size, mainly in the northern part of the area.

Topographically, the Pampa El Toro Dune Field is a perched dune field situated high above the valley floor at elevations of between 1000m and 1400m.a.s.l., rising in elevation from the south towards the north and northeast.

6

History

The Pampa el Toro dune sands were investigated for their iron, titanium and vanadium content by the Marcona Mining Company between 1959 and 1961.  The high titanium and vanadium content of the magnetite, untreatable at that time, ended the project in the early 1960’s.  Later, the Peruvian Ministry of Energy and Mines identified the iron potential of the sands in 1974 as part of a United Nations sponsored national iron inventory study.

The Iron Sands Project was initiated by Cardero Resource Corp. in mid-2005, following the acquisition of magnetite-bearing sand-dune properties located southeast of Nazca.  Preliminary surface sampling of magnetite-bearing active dune sands (a suite of 12 samples from the Carbonera Dune Field), and separation of the magnetic fraction, gave magnetic concentrates (MC’s) in the range 0.7 to 21.24 Wt.%, with an average of around 11.0 Wt.%. Geochemical assays of the magnetic concentrates by ALS Chemex Labs, North Vancouver, B.C., returned results of about 61.5% Fe and 4.3 to 6.4 % TiO₂.

Also during 2005, magnetic concentrates separated by SGS Lakefield Research, Ontario, from two preliminary bulk samples from the Carbonera Dune Field were sent to Midrex Technologies Inc., a division of Kobe Steel, in North Carolina.  An initial Midrex FASTMELT test successfully produced a high-quality low-sulphur, low-phosphorus and low titanium liquid iron of approximately 91.6% iron, 4.88% carbon and <0.21% sulphur (Ripke, 2005).  The FASTMELT^® Process utilizes the same basic flowsheet and equipment as FASTMET^®, but includes an electric melter to produce FASTIRON^®, a high quality liquid iron product. Hot Direct Reduced Iron (DRI) is discharged from the Rotary Hearth Furnace (RHF) and melted in an electric furnace or coal-based melter. The FASTMET^® Process is a solid carbon-based reduction technology using a rotary hearth furnace (www.midrex.com). Waste slag from the process reportedly assayed up to 21.48 Wt.% TiO₂  and 3.18 Wt.% V₂O_5. (Cook, 2006).

Based on these early positive Midrex results, Cardero decided to continue with the Iron Sands Project with the objective of identifying a logistically easily accessible area containing the potential for sufficient resources of magnetite sand at sufficient concentration to support a viable 20 to 30 year integrated mining operation.  To this end, Cardero focussed initial exploration work on the larger Pampa El Toro and Carbonera dune fields, and later over the highest-grade parts of the Pampa El Toro Dune Field.

7

Geological Setting

Regional basement in the coastal area between Marcona and Yauca consists of gneissic and schistose Precambrian metamorphic rocks separated, by angular discontinuity, from silicified and metamorphosed carbonate rocks of the overlying Lower Palaeozoic Marcona Formation.  It is the Marcona Formation which hosts the Marcona Iron Mine, an Iron Ore Copper Gold (IOCG)-style iron oxide deposit of around 1,440Mt at 54.1% Fe (Hawkes et al, 2002).  The Marcona Formation is unconformably overlain by Triassic, Jurassic and Cretaceous volcanic and sedimentary sequences which, in turn, were intruded by stocks and dykes belonging to the Peruvian Coastal Batholith complex.  These strata were, in turn, deformed, uplifted and eroded, and formed a base on which Quaternary sediments were deposited.

The Marcona iron ore has a different chemical signature than the iron sands, which contain elevated Ti and V concentrations.  It is speculated that the iron-bearing minerals contained in the iron sands either partially or wholly originated from the volcanic units overlying the Marcona formation, however additional as yet unidentified sources are possible.

A Cardero consulting geologist, Gary Belik, reported the presence of a Quaternary basin flanking rising ground in the study area, with exposures of friable sand sequences locally containing visible magnetite (Belik, 2005).  However, Belik also noted that most of the Quaternary sections were not magnetite-bearing, and were disrupted by lenses and layers of barren conglomerate, clay and volcanic ash, and concluded that the Quaternary sediments did not present an attractive exploration target for iron, relative to the active dune sands (Cook, 2006).  However, it is possible that erosion of these Quaternary sediments contributed significant quantities of magnetite material contained in the dune fields.

Mineralogically, the principal iron-bearing minerals in the dunes consist of magnetite, titanomagnetite and ilmenite, with minor to trace amounts of hematite, rutile, titanite, ulvospinel and goethite.  Silicate minerals, which make up the bulk of the sands, are primarily quartz, plagioclase, amphibole, chlorite, mica and pyroxene (SGS Lakefield Research, 2005a).  The majority of the ilmenite (80-85% by volume) is associated with magnetite, titanomagnetite and hematite, with the remaining 15-20% by volume present as free ilmenite grains.

The dune sands are considered to be the product of aeolian re-distribution of contemporary and ancient beach sands and Quaternary basin sediments derived from erosion of the bedrock, winnowed, and re-distributed, by prevailing winds (Gay, 2005) to produce potentially economic concentrations of magnetite in extensive, and partially active, dune fields.

8

Deposit Types

The dune sands fall into two deposit types based on dune morphology.  The first is stable pampa sand, consisting of a flat-to-gently-rolling sand apron surface similar to an “erg” or “sand sea”, with small active dunes superimposed.  The second is active dunes with little evidence of a stable sub-stratum.  The deposit may be classified as a heavy minerals deposit, although one generated by wind rather than water processes, as is the norm.

The Pampa El Toro Dune Field is aligned approximately N-S around the western and southern flanks of the Cerro de Arena.  Surface area is >120km², although 2005 sampling was confined to ~63km² in the southern and eastern parts of the field.  Around two thirds of this area (~41km²) is comprised mainly of stable pampa sands, predominantly in the south and south east.  The remaining 1/3 of the area consists of active sand dunes, mainly in the northern parts of the area.  Active sand dunes are relatively small, and of similar size.  Cardero holds claims in ~100 km² of the dune field.

Topographically, the Pampa El Toro Dune Field is a perched dune field, situated high above the desert plain.  Most of the area lies between 1000m and 1400m.a.s.l., elevation increasing from the south towards the northeast.  This is ~1000m above the level of the adjacent Acari Valley (elevation ~150m a.s.l.).  Sand depth is a minimum of 141m over most of the dune filed, and geological reconstruction by Cardero suggests the dune field may attain thickness approaching 300 meters in places, although ultimate depths are unknown.

In general, particle sizes of pampa sands appeared to be somewhat coarser than active sand dunes, perhaps due to active winnowing, although both are fine-grained.

9

Mineralization

The magnetic minerals component of the dune sands is believed to have been derived from erosion of the bedrock, and also later Quaternary basin sediments which probably received their magnetic minerals from the same source.  The magnetic minerals were transported, together with non-magnetic components, by prevailing winds (Gay, 2005), to produce potentially economic concentrations of magnetite in extensive, and still active, dune fields.

Magnetic minerals are distributed throughout the pampa-type surface sands and the active dune sands as rhythmically repetitive millimetre thick stratified bands. The variable intensity of the mineralization, both vertically and horizontally (up to 15.36 Wt% magnetic concentrate) is probably due to differential winnowing of pampa-type and active dune sands by winds of different strengths over different periods and in different locations.

Grain size analysis of sands from Pampa El Toro showed that surface pampa-type sands were generally somewhat coarser-grained than surface sands taken from active dunes in both of those two dune fields.  This may be explained by winnowing of the surface sands in areas with stable pampa-type sands.  

Le Couteur (2005) screened two active dune sand samples from the Pampa El Toro Dune Field, and found that 63.5% (for the coarser grained of the two samples) and 80% (for the finer-grained sample) of the magnetic mineral concentrate, respectively, was contained in the <150 micron fraction, with most of the remaining magnetic fraction reporting to the slightly coarser 150-250 micron range.  Bateman Minerals and Metals (Pty) Ltd., reported an average particle size of 125 micron for MC “sinks” (TBE separation) carried out on a composite of ‘milled’ drillhole samples and consequently is not considered to be representative of the dune field (Rademeyer, 2006).  Deeper, pampa-type sands obtained during percussion drilling were finer-grained than either the active dune sands or the pampa sands, although there is overwhelming evidence (including rock flour in recovered samples) that there was milling of the sands during drilling.

During Phase II percussion drilling and accompanying trenching, it was observed that the grain size of material collected from the drillholes was significantly finer than that encountered during trenching (Torres, 2006a).

The presence of fines in drillhole samples was corroborated by SRK who noted that material was being lost as fines through the cyclone exhaust and that fines captured as part of the sample were problematical when passed through the magnetic separator.  Fines tended to agglomerate and bounce off the drum during separation (Van der Merwe, 2006b).  This raises the possibility that the MCs of drillhole samples are systematically understated due to loss of magnetic minerals to fines.  The most recent work by SRK indicates that these should not create any issues from a resource estimate perspective since the resource estimate is done on an in situ basis, before any magnetic concentration, and loss of fines during the drilling is not considered material because of the small weight of material lost. Furthermore the pilot plant tests conducted in 2008 have shown that magnetic concentrate recoveries from Pampa el Toro are a maximum of 7.5 wt% MC (see section 16.16).

Preliminary mineralogical studies by Le Couteur (2005) indicated that the magnetic concentrate consisted mainly of titanomagnetite (Fe²⁺ Fe³⁺, Ti₂O₄), magnetite (Fe₃O₄) and minor ilmenite (FeTiO₃).  Detailed point counts of the various minerals were not conducted during this study.  However, on the basis of 45 mineral grains in a polished section of the sample, Le Couteur estimated the composition to be around 50% titanomagnetite, 40% magnetite and 10% ilmenite.  It was also noted by Le Couteur that titanomagnetite and magnetite have been partially altered by martitization to titanohematite and hematite grains respectively, and that this alteration not only frequently follows octahedral planes bisecting the grains but also forms grain rims.

10

Exploration

10.1

Surface Sand Sampling

10.1.1

Reconnaissance Surface Sand Sampling Phase

The sampling crew was headed by consulting geochemist Stephen Cook (P. Geo.) of Cook Geochemical Consulting and supplemented by a four-person dune buggy crew of drivers and mechanics. (the two dune buggies working in tandem for safety reasons) Reconnaissance surface sand samples were collected in each dune field at a 1km sample site spacing along north-south lines 500m apart, with samples along adjacent lines being collected at 500m offsets as shown in Figure 9.1. This corresponds to a northwest-southeast / northeast-southwest orientated square grid with a sample spacing of 707m.  The area sampled over pampa sands and dune sands at Pampa El Toro amounted to around 70-75 km²

Regional surface sand samples were processed in Canada at Vancouver Indicator Processors Inc. laboratory, Burnaby, B.C.  Magnetic concentrates were prepared for each sample and weight-per-cent magnetic concentrate (Wt% MC) content calculated.

10.1.2

Infill Surface Sand Sampling Phases I and II: Pampa El Toro

An infill surface sand sampling programme (Phase I Infill Survey) was initiated in mid-January, 2006, over Toro West, broadly covering the extent of sites reporting high (>5.0 Wt.% MC) values which were known from analytical results at that time.  The area covered by Phase I Infill Survey is shown in Figure 9.2.  The objective of the infill surface sampling was to provide a more detailed understanding of grade and MC distribution in the surface sands by increasing the original reconnaissance grid spacing (1000m x 500m, with 500m offset) to one of 250 m x 250 m along a N-S and E-W oriented grid.  A total of 235 sites were sampled, covering an area of approximately 15 km².

Following the infill surface sand sampling at Toro West, a Phase II infill survey commenced in April, 2006, over the smaller Toro Southeast Zone, and also an area adjacent to the southwest corner of Toro West (Figure 9.2).  The distribution of samples, together with those of Phase I Infill Survey, are shown in Figure 9.2.  A total of 168 sites were sampled in Phase II Infill Survey (of which results are available for 164), covering an area of approximately 10 km².  

Figure 10.1: Pampa el Toro reconnaissance surface sand sampling, distribution and results.

Figure 10.2: Pampa el Toro infill surface sand sampling, phases I and II, distribution and results.

10.2

Trenching

The results from Phase I percussion drilling showed a marked fall-off in MC values and grain size of material recovered from the first few meters of the drill holes, downwards.  Therefore, it was decided to trench alongside existing drillhole sites to the maximum depth (5-6 metres) attainable with the equipment available in order to try to resolve this discrepancy.

A total of 41 trenches were excavated variously by bulldozer and excavator (as detailed in Section 12.3.1) adjacent to previously drilled Phase II percussion drillholes.  The objective was to compare Wt.% MC separated from paired drillhole and trench samples over the 0 – 6 m depth interval from surface.

Reconnaissance surface sampling, Phase I infill surface sampling, the Toro West Zone of Phase II infill surface sampling, and Phase I Percussion Drilling Programme were all carried out by contractor Cook Geochemical Consulting, of Victoria, BC, Canada.  

The Toro Southeast Zone of Phase II infill surface sampling, Phase II Percussion Drilling Programme and trenching were carried out by Cardero personnel under the supervision of Cardero’s Project Geologist following the protocols previously established by Cook Geochemical Consulting.

11

Drilling

11.1

Percussion Drilling

11.1.1

Percussion Drilling Programme:  Phase I

A widely-spaced percussion drilling programme was conducted in November and December, 2005, over the Pampa El Toro Dune Field.  Nine drillholes were drilled mainly on the flat-to-gently-rolling pampa sands in the southern and central part of the dune field.  The holes were drilled with nominal 2 km spacing along E-W oriented lines, with drill sites along adjacent lines being offset by 1km.  Most of the drillholes were drilled to a depth of 100m, with two holes (PET-5 and PET-6) drilled to 141m.  Eight of the drillholes terminated in sand and therefore remain open in depth.  However, drillhole PET-12, one of two along the southern edge of the dune field, encountered bedrock at 38 m depth.  All these, and later drillholes, were drilled vertically.

Collar elevations of the drillholes ranged from 1,000 m.a.s.l. in the southern part of the area to a high of 1,529m.a.s.l. in the northeast.  The percussion drilling programme was truck-supported and operated from a base in Lomas.  Drilling services were provided by GeoTech of Lima, equipment including a truck-mounted percussion drilling unit, a second truck with hydraulic crane to remove casings, a D-8 bulldozer to pull the trucks from site to site across the dune field and a local tractor contracted for miscellaneous tasks.

Preparation of magnetic concentrates from these drillholes, and their subsequent assay for %Fe₂O₃, %TiO₂ and related constituents were carried out at Acme Labs in Vancouver.  

11.1.2

Percussion Drilling Programme:  Phase II

Following a recommendation from SRK (Van de Merwe, 2006), a 72 drillhole definition campaign was designed and conducted between May and July 2006, to systematically drill-test an approximately 6.5km² area of the Pampa El Toro dune field to a depth of 60m.  The programme included a combination of holes at both 500m drill centres and 250 meter drill centres (Error! Reference source not found.).  Approximately 2.5 km² of the pampa sands were tested by holes with 250m centre separations.  

The definition drill program was designed to test the central part of Toro West Zone where Phase II surface sand sampling had defined a broad area containing >5.0 Wt.% MC within which were two higher-grade areas, shown as Zone “A” and Zone ”B” in Error! Reference source not found., reporting mean values of 8.6 and 9.7 Wt.% MC respectively. Additional holes were drilled at the end of the programme to capture parts of zones “C” and “D” (Error! Reference source not found.).  All of the drillholes were targeted to 60m.

Following the completion of the 72 drillhole definition campaign, a repeat of Phase I drillhole PET-08 was drilled 5 m from the original in order that the results might be compared between Phase I and Phase II drilling.  It was during the process of splitting the samples from the PET-08 repeat hole that it became clear that the material retrieved appeared to be much finer grained than the surface sands with a significant amount of dust-particle-size fines, or “rock flour”.

At the drilling company’s suggestion, and in order to check whether or not the percussion bit in use was grinding the sand during drilling, a different type of bit with characteristics that might not generate the same amount of in-hole friction was employed for the remainder of the Phase II drilling programme.  This included a) the additional 10 drillholes capturing parts of zones “C” and “D”, mentioned above, b) duplicates (or “twins”) of 10 Phase II drillholes already drilled and c) a further 10 drillholes (an “L” set), to provide details for geostatistical analysis, also on the recommendation- of SRK Consultants (Figure 11.2).  The characteristics of the samples appeared to be unchanged.

Field and laboratory processing of Phase II drillhole samples was carried out under the supervision of Cardero’s Project geologist.  Representative splits of borehole sand samples were processed at Cardero’s on-site magnetic separation laboratory at Lomas.

Figure 11.1: Map of percussion drill hole collars.

Figure 11.2: Map of percussion drillhole duplicate and trench locations.

12

Sampling Method and Approach

12.1

Reconnaissance and infill surface sand sampling

12.1.1

Field collection methods

In all cases, field sampling equipment consisted of a Jones riffle splitter with stainless steel catch pans, a series of steel sample collection pans, and a short-handled shovel.

Two different sampling techniques were employed for the pampa and active dune sands respectively.

a)

Pampa sands

After digging an 80-90 cm deep pit in the relatively flat-lying pampa, three vertical channel samples were taken by sweeping a shovel blade upwards from base to top along different parts of one sidewall of the pit (Figure 12.1).  Each of the three channel samples was split once in a Jones splitter, and a composite sample made by combining one of each of the three splits.  The remaining halves of the three splits were discarded.  The final sample weighed 10-12 kg.

Figure 12.1: Schematic diagram showing field procedure for collecting reconnaissance sand samples at pampa sand sites.

b)

Active dune sands

A different collection method was adopted for the active dune sands, where a greater variation in results was considered possible from the constantly-shifting dunes.  Nine individual sub-samples were collected on a north-oriented cross, with the designated sample site location at its centre (Figure 12.2).  Each of the nine sub-samples were taken as one “spade-full”, after scraping away the upper several centimetres of surface sand.  These individual samples were then riffle-split twice, and composited to produce a final sample weighing around 10-12 kg.

Figure 12.2: Schematic diagram showing field procedure for collecting reconnaissance sand samples at dune sand sites.

12.1.2

Sample documentation and security

Sample site locations (UTM coordinates, WGS 84 datum, 18S zone) were selected in advance to ensure that any sample site selection biases were eliminated.  Sites were located by pre-entering coordinates into a Garmin GPS and utilizing the “go-to” feature to travel to the spot with a horizontal accuracy of approximately 5m.  Only a few sites were moved, minimally, from their pre-ordained locations as a result of local conditions such as steep dune slopes.

At each site, UTM coordinates were re-read and recorded on field cards, along with sample number, site number, sample-type information, site elevation, duplicate / standard status (if applicable), reference to digital photo of sample site and other relevant data as required.  Sample numbers were written two or three times onto heavy-duty polythene sample bags using indelible markers, and a numbered ticket detached from the respective field card placed in each bag.  Each sample bag was secured with heavy string and placed within protective rice sacks for transport via dune buggy and truck to the field camp at Lomas.  

All field samples were temporarily stored in a locked warehouse with night-watchman cover in Lomas, prior to being double bagged, trucked to Lima, and sea-freighted to Vancouver.

12.1.3

Quality control procedures in the field

Quality control procedures used to help quantify the extent of site and sampling variations included the collection of field duplicate samples, within-pit duplicate samples and splitter (check) duplicates at regular intervals, and also the incorporation of field standards into the collection suite.  

Field duplicates were collected at the rate of one pair every ten samples.  Pampa site duplicates were collected by the method noted above (11.1.1a) from two adjacent pits dug approximately 3.5m apart.  Dune sand site duplicates were collected by the method noted above (11.1.1b) but by moving the sites of individual sub-samples (and therefore the north-oriented cross as a whole), 3 metres to the east and 1 m to the north to create a new identical sampling cross with individual sub-sample separation equivalent to that between pampa sample duplicates (Figure 12.3).

Within-pit duplicates also were collected at the rate of one pair every ten samples.  Pampa sands within-pit duplicates were taken by the same methodology from the opposite wall of the same pit. Dune sands “within-pit” duplicates were produced by repeating the process described in 11.1.1b above with a second sub-sample “spade-full” taken immediately adjacent to the first (Figure 12.4).

Splitter duplicates also were collected at the rate of one pair for each ten samples, and involved bagging separately but numbering identically each of the respective splits for a given sample site.  However, the practice of taking riffle splitter duplicates to test for variations introduced by the splitting method was later discontinued in the surface sampling programme as no significant differences in results were detected between splitter duplicate pairs.

Slots for insertion of field standard samples, included to monitor magnetic concentrate separation processing accuracy and precision, were left open every 20 samples. As no standards are commercially available for the separation of magnetic minerals from iron sands, a field standard was prepared from a high-magnetite site at Pampa El Toro. In the analytical assay suites, the magnetic concentrate from this sample was used as a drift monitor.

Figure 12.3: Schematic diagram showing procedure for collecting field duplicate samples during reconnaissance sand sampling at dune sand sites.

Figure 12.4: Schematic diagram showing procedure for collecting within-pit duplicate samples during reconnaissance sand sampling at dune sand sites.

Figure 12.5: Typical 20-sample collection “block” used in The Pampa El Toro reconnaissance surface sands sampling programme.

A typical 20-sample collection block used in the Pampa El Toro surface sands sampling program, showing duplicate, field standard insertion frequency and labelling protocol is shown in Figure 12.5

12.2

Percussion Drilling

12.2.1

Field collection methods

Sand samples were obtained from the drill rig at one metre intervals, discharged into heavy polythene bags, tied, labelled, and left onsite for splitting.  Each 1 metre sample weighed approximately 40-45kg.  Five-metre composite samples were prepared by splitting the sand recovered from each interval using a large “Jones Riffle Splitter”.  In each case, composites were prepared by pouring the five bags (approximately 200-250kg total) onto large plastic trays, homogenizing, and splitting three times to produce a single 25-30kg composite sand sample for each 5m interval.  These samples were transported to the Lomas laboratory for magnetic separation.  Reject drilling samples are stored at the drilling site, buried and covered with sand to prevent deterioration of the bags due to ultraviolet radiation.

In 2008, the reject drilling samples were recovered, re-split, employing the same procedures, and reduced to 5kg representative composite samples for each 5m interval of drilling.  These samples were transported to the laboratory in Lomas and are stored as representative raw (unseparated) sand samples from the drilling campaigns.

12.2.2

Sample documentation and security

Sample documentation and security followed the same protocol as that adopted for reconnaissance and infill surface sand sampling described above. However, differential rather than hand-held GPS was used to position holes, and later to check coordinates of drill sites with greater confidence.

12.2.3

Quality control procedures in the field

Quality control procedures included the collection of splitter duplicates at regular intervals and the incorporation of field standards (for magnetic separation) and certified reference materials (for analytical quality control) into the sample suite. Given the limitations of collecting drill hole samples compared to surface samples, it was not possible to collect the equivalent of field duplicate or within-pit duplicate samples. It should be noted that in this case (cf. surface samples) the numbering scheme used for splitter duplicates ensures that their relation to adjacent samples remains blind to the processing laboratory (Figure 12.6).

Splitter duplicates were collected at a rate of one pair every 10 samples for the magnetic separation work, and one pair every 20 samples for the analytical work.  The location of splitter duplicates in the drill hole were chosen in advance, and the duplicates were prepared by retaining both halves of the third, and final, split of these samples during riffle splitting.  In these cases the usual split was retained for the routine sample, and the other split was retained as the splitter duplicate sample.  

In the case of the raw (unseparated) sand samples, analytical (pulp) duplicates were also inserted, by means of leaving an ‘open’ sample number (empty bag) every 1 in 20 samples. This alerted the laboratory to take a split of the previous sample after the pulverization stage.

Slots for insertion of a control standard were left open at a rate of one in 20 drilling samples. A thoroughly-homogenized field standard was inserted for monitoring the quality of the magnetic separation work and as an analytical drift monitor in magnetic concentrate analyses. Certified reference materials were inserted to monitor analytical accuracy in the raw (unseparated) sand samples. An internal standard was created (MAG-1) from pulverized and homogenized magnetic concentrate material and inserted into the magnetic concentrate analytical sample batches.

A typical 20-sample collection block used in drill hole sampling, showing splitter duplicate, control standard insertion frequency and labelling protocol is shown in Figure 12.6. In the case of Phase II drilling at Pampa El Toro, where the holes were drilled to 60 m depth, it was ensured that a single pair of splitter duplicates and a single field standard were included in each drillhole.

Figure 12.6: Typical quality control scheme used in the sampling of sands from a typical 100 metre-deep percussion borehole at Pampa El Toro.

12.3

Trenching

12.3.1

Field collection methods

Trenching methods included the use of both a Cat D-8 Bulldozer, when this was on site to assist with the percussion drilling phase, and a Cat 330C Excavator.  Trenches were dug immediately adjacent to actual borehole sites for direct comparison purposes, and care was taken to avoid areas where the pampa surface had been disturbed by drilling-related activity or contaminated by drill-generated fines.  In the case of the bulldozer, four samples were taken in the same way as pampa sands (section 11.1.1a, above), with the first metre being sampled via an 80-90 cm deep pit. The first metre of sand was then scraped off with the bulldozer blade, using a graduated pole as depth reference, to form a platform from which the next pit could be dug to sample the 1.0-2.0 m interval.  In this way it was possible to sample 1.0 m intervals to a depth of 4.0m to 5.0m (Torres, 2006a; Torres, 2006b).

In the case of the 37samples obtained by excavator, the first sample from surface to a depth of 1.0 m was collected with a single rotation of the excavator bucket.  This was unloaded well away from the sides of the designed trench to form a small pile, from which a representative sample of around 25 kg was then taken and bagged. The surrounding sand was then excavated to a depth of 1.0m over an area sufficient to ensure that the walls of the trench did not collapse into the trench and contaminate subsequent samples as the trench was deepened.  The second sample was then collected with a single rotation of the excavator bucket between 1.0m and 2.0m. This process was repeated in 1.0m intervals to 5.0m or 6.0m, depending on operating conditions.

12.3.2

Sample documentation and security

Sample documentation and security followed the same protocol as that adopted for reconnaissance and infill surface sand sampling described above. Samples were numbered in a way which related them to the adjacent borehole and depth from which they were collected.

12.3.3

Quality control procedures in the field

Trench sites were located adjacent to existing percussion drillhole sites in order to verify the consistency of data obtained by percussion drilling over the first five metres. Due to the instability of loose, dry sand under trenching conditions, and the probability of “run-in” of loose material and contamination of repeat samples taken from the same point, field duplicate and within-pit duplicate samples were not included in the sample suite. Splitter duplicates were also omitted because of the lack of significant differences detected between splitter duplicate pairs in the results from reconnaissance and infill surface sand sampling.

12.4

Bulk Sample Collection

12.4.1

Field Collection Methods

A 2500 ton bulk sand sample from the Pampa el Toro property was collected in November-December 2007 for processing at a magnetic separation pilot plant facility in the town of Acari.  The sample was taken from two separate areas within the high-grade core of the property to a depth of 1m with the use of a backhoe and transported by truck to a secure pilot plant site facility in Acari.

The sample is determined to be representative of the known magnetic concentrate grade over the high-grade core of the property where contingent on ongoing results any future mining operation is most likely to commence; however, it was not possible to adequately test the sub-surface below 1m depth.

12.4.2

Sample documentation and security

Tonnage by truck haul was tallied at the site and 2500 tons were collected.  The stockpile was covered securely in tarpaulin to avoid contamination and/or upgrade of the bulk sample due to aeolian processes.  The site facility is a secure area with guarded surveillance twenty-four hours a day.

12.4.3

Quality control procedures in the field

The sample areas were located by GPS, staked and flagged in the field.

13

Sample Preparation, Analyses and Security

13.1

Magnetic Separation and processing

13.1.1

Magnetic separation and processing of reconnaissance surface sand samples

All reconnaissance surface sand samples were trucked from the Lomas field warehouse to Lima, Peru and then shipped to North America.  Samples were shipped in secure wooden crates to protect against any damage to the plastic sample bags while in transit.  They were forwarded to Vancouver Indicator Processors Inc. laboratory (VIPI), Burnaby, B.C., Canada, for magnetic separation of magnetite grains from the sands, and subsequent determination of the percent magnetic concentrate in each sample.

VIPI laboratory was primarily a heavy mineral lab for the diamond exploration industry, and was equipped with suitable magnetic concentration equipment for processing iron sands samples.  All magnetic separation work was conducted by or under the direction of Dr. Peter Le Couteur of VIPI, and the following details of the equipment and the separation procedure are taken from his report (Le Couteur, 2005).  

The field samples, generally between 9kg and 13kg each in weight, were weighed as received at the Burnaby laboratory, and magnetic concentrates prepared using a ‘scalper’ magnet to remove only ferromagnetic mineral grains (ie. only magnetic, not paramagnetic, mineral particles).  The scalper unit comprises an Eriez sample hopper and vibratory feed chute coupled with a new Outukumpu Model LP10 dry permanent magnet with a weak magnetic field (approx. 0.05 Tesla).  The sample was fed from the hopper down the vibrating chute, and slides under gravity across the magnet.  The magnet attracts ferromagnetic particles into one receiving pan beneath the magnet, allowing all other non-magnetic grains to pass undeflected into a second receiving pan.  A single-pass procedure was used for all sand samples.

13.1.2

Magnetic separation and processing of drillhole and infill surface sand samples

Magnetic separation of drillhole and infill surface sand samples from the Toro West and Toro Southeast infill surveys, was carried out on-site at Cardero’s Lomas field facility beginning in January, 2006, using an Eriez Model FR ferrite radial drum magnetic separator.  This work was carried out by, or under the direction of, Cardero’s Project Geologist.  

The equipment and laboratory was tested by Dr. P. Le Couteur in January, 2006, and set up for optimum performance on-site (Le Couteur, 2006).

Drillhole and/or infill surface sand samples were weighed in the warehouse using a high-capacity Bascula PCR-40 electronic balance (40 kg capacity), and then processed using the magnetic separator.  Material was fed from the sample hopper onto the rotating magnetic drum using a vibratory chute feeder.  Non-magnetic reject material passes over the rotating drum and was captured by a large tray prior to re-bagging and storage.  Ferromagnetic minerals attracted to the drum were captured in a smaller tray positioned beneath it.  After primary magnetic separation, the magnetic concentrate was fed through the separator a second time to remove any adhering silicate grains.  The subsequent, second-pass concentrate was retained, bagged and weighed using a Bascula LEQ-5 electronic balance (5kg capacity), and the weight per cent magnetic concentrate content calculated. The weight of the first-pass concentrate was also recorded, and the resulting ‘middlings’ sample fraction from the second pass bagged and retained for any possible future use (Cook, 2006).

Following observations by SRK during an April 2006 visit to the Lomas field facility (Van der Merwe, 2006b), Cardero field operators introduced a third pass to the separation process whereby the second pass concentrate was fed through the separator a third time to remove remaining non-magnetic material and to further clean the concentrate.  Van der Merwe had noted that a comparison of four samples that were processed by both the Vancouver Indicator Processors Inc. and the Lomas facility showed consistent better recovery of ‘magnetic concentrate’ by the Lomas separator.  It is possible that the Lomas machine might have been over-recovering “magnetic concentrate” by including some non-magnetic fraction (Van der Merwe, 2006).  Therefore, from April 2006, it was the third-pass concentrate that was retained, bagged and weighed, and the weight per cent magnetic concentrate content calculated.  The weight of the second-pass concentrate was also recorded, and rejects from the third pass were added to the ‘middlings’ sample fraction from the second pass noted above.  

13.1.3

Quality control and security

a)

VIPI Laboratory

Le Couteur (2005) conducted several tests to determine the efficiency of magnetic separation at the VIPI Laboratory.  Results of these tests suggested that the VIPI scalping magnet recovers most (approximately 90% of the total ferromagnetic material) in the first pass.  Tests of several samples which were processed using between four passes and ten passes through the magnetic separator showed that most of the magnetite, about 87-94% of the total, was removed in the first pass.  Typically the second pass yielded only a small amount of magnetite, in the range 3.3–4.7%, and subsequent passes yielded progressively smaller and smaller amounts.  Nevertheless the single sample which was subjected to ten passes through the separator (87% magnetic mineral recovery on the first pass) yielded a small amount of magnetic material even on the tenth pass (0.6%).

Reproducibility testing, involving two samples which were processed 5 times each, showed a high degree of replicate precision in the magnetic separation results.  Replicate results were within about ±3% of mean values, varying only narrowly in the range 4.11-4.21% in one case, and in the range 1.77-1.87% in the second case (Le Couteur, 2005).

b)

Lomas Facility

Following initial set up of the magnetic separator at the Lomas field facility in December, 2005, testing of optimal operating parameters for processing iron sands samples was carried out VIPI Laboratory during a visit to Lomas in mid-January 2006 (Le Couteur, 2006).  Le Couteur determined that an extraction efficiency of about 98% magnetite was achieved from test samples using a feed rate of 3.5 kg/minute and a drum rotation speed of about 118 rpm.  These settings were adopted as the standard separator operation settings for the Pampa El Toro sand samples although, as he notes in his report, a wider range of feed rate settings is possible.  Tests showed that a 25kg sand sample could, ideally, be processed in about 7 minutes at the recommended settings.

A calibration standard, prepared by Le Couteur, was run at the start of every magnetic separation processing session in Lomas in order to test the functioning of the magnets and the apparatus in general.  This calibration standard is separated, the weights are recorded, and the standard is subsequently re-combined for processing at the start of the next session.  The reproducibility of the results has shown to be exemplary, and this demonstrates that any over-time-drift of the magnet and/or apparatus function is insignificant.

13.2

Sample analysis and assay procedures

Raw (unseparated) sand samples from all boreholes (5 m intervals) were submitted for analysis. In addition, a subset of magnetic concentrates and ‘middlings’ samples were analysed at Acme Analytical Laboratory (ISO 9001:2000 accredited), Vancouver, British Columbia. These samples form the basis of the data that was used for Mineral Resource estimation.

13.2.1

Sample Preparation

For raw (unseparated) sand analyses, 250-300 g of material are riffle-split from the representative drilling interval samples (5kg each) at the Lomas facility by Cardero employees. The samples are forwarded to Acme Labs for pulverization by ceramic ring mill and analysis.

Upon arrival at the Acme Analytical Laboratory, Vancouver, each magnetic concentrate sample is rolled thoroughly on a large clean piece of paper to ensure that it was adequately mixed prior to the separation of an approximately 200-250 g sub-sample.  The sample is pulverized, using a ceramic, rather than steel, ring mill to prevent any possible iron contribution to (and contamination of) the concentrate during ring milling.  All unused magnetic concentrate material are retained in the original sample bag; and the pulverized magnetic concentrate pulp material is retained in an additional bag.

13.2.2

Analysis

All samples were analyzed at Acme laboratories for potentially economic elements including total iron, titanium and vanadium, and for potentially deleterious constituents such as %SiO₂, %Al₂O₃, %MgO and %P₂O₅, by the following total-decomposition methods:

a)

Group 4A major element oxide suite

This whole rock package uses a LiBO₂ fusion and dilute nitric acid digestion of a 0.2 g sub-sample with an Inductively Coupled Plasma – Optical Emission Spectroscopy (ICP-OES) finish.  Constituents determined are SiO₂, Al₂O₃, Fe₂O₃, CaO, MgO, Na₂O, K₂O, MnO, TiO₂, P₂O₅, Cr₂O₃, LOI, Total C and Total S.  The analytical upper limits of each of these determinations are stated by Acme Labs to be 100%, suitable for the high-iron composition of the magnetic concentrates.

b)

Total vanadium

Determination of total vanadium was conducted as part of the Group 4B suite of elements, including the rare earths.  This total determination method involves the fusion of a 0.2 gram sub-sample by LiBO₂/Li₂B₄O₇ fusion followed by an ICP/MS finish.

The magnetic concentrate samples submitted from the Phase II drilling program followed the same analytical procedures as listed above, however with a 0.1 g sub-sample weight (refer to section 14.1.2).

14

Data Verification

14.1

Cardero Verification

14.1.1

Quality Control Procedures

Magnetic concentrates of all duplicate samples including field, within-pit and splitter duplicates for reconnaissance and infill surface sand samples, and splitter duplicates from Phase I drillhole sand samples were analyzed along with concentrates from the routine surface sand and drillhole sand samples.  The field standard sample was analyzed in replicate seven times for the reconnaissance suite.

CANMET iron ore CRM MW-1 was initially inserted as an assay standard in the reconnaissance program, but after the laboratory reported continued problems in fusing this material, two drift monitors were instead created from the pulverized pulp material of two routine borehole samples.  Multiple insertions of the two sets of drift monitors into the magnetic concentrate assay suites permitted an evaluation of analytical precision, if not absolute accuracy.  Raw sand borehole samples included splitter duplicates and preparation duplicates, each inserted at a rate of one per twenty samples.  Preparation duplicates were split at Acme laboratories from the original sample after pulverization and before analysis.  A Certified Reference Material (OREAS 42P, 8.8% Fe; GBAP6, 4.8% Fe; or Blank, <1.5% Fe) was also inserted into the sample sequence at a rate of 5%.

The laboratory also reports their internal quality control data, which includes pulp duplicates (identical split as the preparation duplicates), blanks and standards each at a rate of approximately per forty samples.

14.1.2

Internal and External Check Assays

Subsequent to the primary magnetic concentrate assays, a suite of magnetic concentrates from two complete drillholes (PET-03 and PET-08) was submitted in entirety for both internal and external check assays, providing a total of three comparative assay results for magnetic concentrates prepared from these borehole sands.  Each hole comprised 20 sample intervals, and in addition to this all splitter duplicate samples and drift monitor insertions which had previously been part of this group of samples were also included in the subsequent check assays.  Internal check assays for iron, titanium and vanadium using the Group 7PF sodium peroxide fusion method for refractory ores was conducted at Acme Labs, using pulverized pulp material which had been previously prepared for the primary assays.  The Group 7PF peroxide fusion method was chosen here to i) provide alternative results to the Group 4A whole rock fusion method which, as all analyses necessarily total to 100%, might be affected by mathematical closure, and ii) to determine if a more aggressive fusion method is required to deal with any refractory minerals which might be present in the magnetic concentrates.  

In addition, external check assays of all magnetic concentrate samples from the same two complete drillholes (PET-03 and PET-08) were carried out at ALS Chemex Laboratories (ISO 9001:2000 accredited), North Vancouver, B.C.  Pulverized pulps were forwarded to ALS Chemex from Acme Labs, and a suite of CANMET CRM MW-1 samples were also sent for inclusion in the ALS Chemex analytical suite as blind standards (Cook, 2006).

The internal and external check assays of the magnetic concentrate samples highlighted an issue with the laboratory initially underestimating the grade of %V and %TiO₂ in the Group 4A and 4B analyses.  Upon further investigation, it was discovered by the laboratory that halving the sample weight would ensure a complete fusion of the sample material.  The results, expressed here as percent change from the original method, are an average increase by 35% (±13%) in TiO₂, 40% (±19%) in V and a decrease by 2% (±2%) in Fe₂O₃.  These results suggest that the original Phase I drillhole analytical results as reported by Cook (2006) may be upgraded to 4.8 wt.% TiO₂ and 2217 ppm V, while Fe₂O₃ is downgraded to 58.0 wt.%.  Given the minor changes in assay grades, Cardero elected not to have the analytical batches re-analysed in entirety.

In the SRK’s opinion, and on the basis of all the information that is available to them, quality control of sample preparation, analytical and assay procedures, and security including chain of custody procedures, followed industry best practice standards.

14.2

SRK Verification

Cardero supplied SRK with assay results from the analysis of two Certified Reference Materials (CRMs) (fourteen results for OREAS 42P and eleven results for GBAP-6) and three sets of samples that were analysed in duplicate by Acme Analytical Laboratories in Vancouver, British Columbia. Forty one duplicates (splitter or field duplicates) were taken in the field during the sample splitting process from the final split. Forty one duplicates identified and requested by Cardero were taken after pulverization (“Preparation duplicates”), and twenty seven pulp duplicates (“Pulp duplicates”) were selected by the laboratory for their own internal quality control after pulverization and at the sample weighing stage. The CRM’s are commercially available pulps created by Ore Research & Exploration Pty (Ltd) for OREAS 42P and Geostats Pty (Ltd) for GPAB-6.

OREAS 42P is a composite standard produced from a range of oxidized materials, but predominantly a gold bearing greywacke. GPAB-6 is a Bauxite pulp sourced from material in the Gove Peninsula, Northern Territory, Australia. Both CRM’s are certified for Fe₂O₃ content, and other major element oxides. OREAS42P was analysed by Borate fusion disc X-Ray Fluorescence (“XRF”), while GPAB-6 was analysed by both XRF and Inductively Coupled Plasma (“ICP”) analyses (the certification is not specific as to the finish used with the ICP method, but this is typically assumed to be Optical Emission Spectroscopy (“OES”)).

Both the CRM’s and the duplicate analytical results have been analysed by SRK and selected statistics and graphics, and the conclusions drawn from the analysis are summarized below. The certified values for selected variable are presented in Table 14.1 and Table 14.2 along with the estimated acceptance limits (calculated as 5% above and below the certified value) and the mean of all the assays and the difference between the average of the assays and the certified value, expressed as a percentage of the certified value.

Table 14.1: Certified values of selected variables for OREAS 42P and statistics of assay results

Variable	Certified Value	-5%	+5%	Mean of Assays	Difference (%)
Fe2O3 (%)	12.60	11.97	13.23	12.48	-1%
SiO2 (%)	62.74	59.60	65.88	62.04	-1%
TiO2 (%)	0.62	0.59	0.65	0.60	-3%
P2O5 (%)	0.09	0.08	0.06	0.09	7%
Na2O (%)	0.19	0.18	1.12	0.17	-11%
K2O (%)	2.83	2.69	0.00	2.88	2%
MnO (%)	0.06	0.05	0.20	0.05	-11%
MgO (%)	1.07	1.02	2.97	1.05	-2%

Table 14.2:

Certified values of selected variables for GPAB-6 and statistics of assay results

Variable	Certified Value	-5%	+5%	Mean of Assays	Difference (%)
Fe2O3 (%)	6.80	6.46	7.14	6.68	-2%
SiO2 (%)	4.41	4.19	4.63	4.45	1%
TiO2 (%)	3.54	3.36	3.72	3.63	3%
P2O5 (%)	0.05	0.05	0.06	0.05	-7%
V (ppm)	173.73	165.04	182.42	190.18	9%

For Fe₂O₃, SiO₂, K₂O and TiO₂ percentages the analyses of both CRM’s show a small difference between the average of the assay values and the certified value. All but one Fe₂O₃ assays fall within the 5% acceptance bracket values, while two of each of the SiO₂ and TiO₂ analyses of the GPAB-6 CRM are outside the 5% bracket values. One of the OREAS 42P K₂O values falls outside the 5% bracket values.

In contrast, of the nine GPAB-6 assay results for Vppm, only one is within the bracket values and there appears to be a small positive bias in the Vppm values. %MgO shows a small negative bias but good precision with none of the assay results falling outside the 5% bracket values. %MnO, %P₂O₅ and %Na₂O are all close to the detection limit. All %MnO values are reported at detection limit, while the P₂O₅ and Na₂O results are highly variable, and are not considered precise at the CRM certified value levels. Time sequence control plots of Fe₂O₃ %, SiO₂ %, TiO₂ % and V ppm are presented in Figure 14.1 to Figure 14.4.

Figure 14.1: Time sequence control plot of analytical results of OREAS 42P and GPAB-6 for Fe₂O₃ %

Figure 14.2: Time sequence control plot of analytical results of OREAS 42P and GPAB-6 for SiO₂ %

Figure 14.3: Time sequence control plot of analytical results of OREAS 42P and GPAB-6 for TiO₂ %

Figure 14.4: Time sequence control plot of analytical results of GPAB-6 for V ppm

Duplicate analyses were performed on a three sets of samples split in the field or after pulping (Splitter, preparation and pulp duplicates) stages of the sample preparation process as described above. Typically the splitter duplicates are expected to have a lower level of accuracy than the pulp duplicates, as the inhomogeneity in the raw samples results in differences between the splits. In this instance, the field duplicates show similar levels of accuracy to the pulp duplicates as illustrated in the three scatter plots in Figure 14.5.  The graphs plot the original assay against the duplicate assay for each of the three datasets. The green line represents where the ideal correlation, where original and duplicate analyses return identical values, the dashed black lines are a 10% error around the ideal correlation, the blue line represents the Reduced Major Axis (RMA) trend line for the data, and the red points represent actual assay results. All three graphs show similar ranges of data, clustering of the data around the ideal correlation line, and RMA lines. The slope, Y axis intercept and errors on the slope and Y axis intercept listed on each graph relate to the RMA line.

Figure 14.5: Scatter plots of Fe₂O₃ duplicate analyses from Splitter, Preparation and Pulp datasets

SRK’s analysis of the duplicate assay results focussed principally on Fe₂O₃, %TiO₂, Vppm and %SiO₂ results, although additional checks were performed on minor constituent element results such as %CaO, %MgO, %Na₂O, %K₂O, %P₂O₅ and %Al₂O₃. In all cases, for all three datasets the duplicate results were found to have a good correlation, low scatter and no evidence of any bias. The correlations and relative errors are consistent for all three datasets as discussed above, and SRK have elected to present some of the analytical results from the Pulp duplicate analytical dataset in this section. Summary statistics for the major element results from the pulp duplicate analyses are presented in Table 14.3. All the variables presented in Table 14.3 have less than 1% difference in the means, and excellent correlation coefficients. The scatter plots of the original versus duplicate pulp analysis datasets for %SiO₂, %Fe₂O₃, %TiO₂ and Vppm are presented in Figure 14.6 to Figure 14.9.

Table 14.3: Descriptive statistics of the duplicate datasets for a range of analysed variables.

Variable	Statistic	Original Assay	Duplicate Assay	Variable	Statistic	Original Assay	Duplicate Assay
SiO2	Count	27	27	MgO	Count	26	26
0.09% *	Arithmetic Mean	64.2	64.3	-0.47% *	Arithmetic Mean	1.70	1.69
0.942 **	Minimum	59.3	59.6	0.996 **	Minimum	1.30	1.28
	Maximum	67.3	67.5		Maximum	2.52	2.52
	Standard Deviation	1.81	1.72		Standard Deviation	0.26	0.26
	Coefficient of Variation	0.03	0.03		Coefficient of Variation	0.15	0.15
Fe2O3	Count	27	27	CaO	Count	27	27
-0.24% *	Arithmetic Mean	6.33	6.31	0.00% *	Arithmetic Mean	4.67	4.67
0.983 **	Minimum	3.87	3.77	0.952 **	Minimum	4.28	4.23
	Maximum	10.21	10.04		Maximum	5.56	5.50
	Standard Deviation	1.71	1.62		Standard Deviation	0.25	0.25
	Coefficient of Variation	0.27	0.26		Coefficient of Variation	0.05	0.05
TiO2	Count	27	27	Na20	Count	27	27
-0.73% *	Arithmetic Mean	0.69	0.68	0.00% *	Arithmetic Mean	3.49	3.49
0.968 **	Minimum	0.45	0.46	0.947 **	Minimum	3.14	3.10
	Maximum	1.10	1.13		Maximum	3.83	3.91
	Standard Deviation	0.17	0.16		Standard Deviation	0.17	0.19
	Coefficient of Variation	0.24	0.23		Coefficient of Variation	0.05	0.06
P2O5	Count	27	27	K2O	Count	27	27
0.75% *	Arithmetic Mean	0.13	0.13	-0.15% *	Arithmetic Mean	2.00	2.00
0.742 **	Minimum	0.11	0.11	0.914 **	Minimum	1.73	1.72
	Maximum	0.16	0.17		Maximum	2.22	2.23
	Standard Deviation	0.01	0.01		Standard Deviation	0.11	0.11
	Coefficient of Variation	0.10	0.11		Coefficient of Variation	0.06	0.05
V	Count	27	27	MnO	Count	27	27
-0.94% *	Arithmetic Mean	162.3	160.8	0.00% *	Arithmetic Mean	0.09	0.09
0.987 **	Minimum	92.0	86.0	0.978 **	Minimum	0.06	0.06
	Maximum	274.0	274.0		Maximum	0.13	0.14
	Standard Deviation	51.13	47.48		Standard Deviation	0.02	0.02
	Coefficient of Variation	0.31	0.30		Coefficient of Variation	0.21	0.21

* The difference between the means of the original and duplicate dataset expressed as a percentage of the mean of the original samples

** The correlation coefficient (R²) of the original and duplicate datasets.

Figure 14.6: Scatter Plot of SiO₂ original and duplicate pulp analyses

Figure 14.7: Scatter Plot of Fe₂O₃ original and duplicate pulp analyses

Figure 14.8: Scatter Plot of TiO₂ original and duplicate pulp analyses

Figure 14.9: Scatter Plot of V original and duplicate pulp analyses

There is no indication of a material bias in any of the datasets, and all but one data point fall within the 10% brackets plotted on the graphs. SRK commonly benchmark the accuracy of duplicate analytical datasets using the Half Absolute Relative Difference (“HARD”) statistics and plots. HARD Plots for SiO₂, Fe₂O₃, TiO₂ and V are presented in Figure 14.10 to Figure 14.13. The HARD value is calculated by the following formula (with A and B representing the original and duplicate analyses)    i.e. the absolute difference of the duplicate assay pair divided by the average of the pair, and then multiplied by 0.5. The HARD value is expressed as a percentage. In a high quality dataset, 90% of the sample pairs have a HARD value of less than 10%. In the HARD plots in Figure 14.10 to Figure 14.13 the target accuracy is represented by the green block. The blue line represents the ranked HARD values for all assay pairs. In all cases the blue line passes through the green block, indicating a high accuracy dataset. The HARD statistics are comparable for all three datasets, and for all the minor elements analysed.

Figure 14.10: HARD Plot of SiO₂ duplicate analyses from the pulp dataset

Figure 14.11: HARD Plot of Fe₂O₃ duplicate analyses from the pulp dataset

Figure 14.12: HARD Plot of TiO₂ duplicate analyses from the pulp dataset

Figure 14.13: HARD Plot of V duplicate analyses from the pulp dataset

In summary, SRK consider that Cardero have a suitable quality control program in place for ensuring the precision and accuracy of their assays, and that the assay results are of sufficient accuracy and precision for use in Mineral Resource estimation. The quality of the assay results does not represent a limiting constraint in the classification of the Mineral Resources.

15

Adjacent Properties

To the knowledge of SRK and Cardero, there are no properties immediately adjacent to the Pampa el Toro property currently undergoing exploration by another company. Cardero Resource Corp. owns the rights to the Carbonera Iron Sands property approximately 50km to the northwest of Pampa el Toro (4,195.12 ha). It underwent initial exploratory investigations concurrently with the Pampa el Toro property. Surface sand samples were collected and anomalous areas identified, however the property has not been drill tested. One bulk sand sample from this property was processed at the magnetic separation pilot plant for investigative purposes.

16

Mineral Processing and Metallurgical Testing

Cardero has undertaken metallurgical test work on block samples obtained from the Pampa el Toro dune site, and the results are described in this section. SRK has not currently undertaken any review of the metallurgical results presented herein. Review of these results is planned to follow the Mineral Resource estimate when SRK have been requested to undertake a Preliminary Economic Assessment on the Pampa el Toro project.

16.1

Mineral Processing and Metallurgical Testing 2005 - 2007

Five bulk sand samples have been submitted to various consultants and laboratories in order to ascertain the optimum process for magnetic concentration of the iron sands, as well as upgrading the recovery of Fe, and in the case of Midrex, %TiO₂ and %V₂O₅ as well.  A summary of the sample origin and results are presented in Table 16.1

Summary of source and results of tests conducted on five bulk samples from Pampa el Toro and two bulk samples from Carbonera – 2005-2007.

Table 16.1: Summary of source and results of tests conducted on five bulk samples from Pampa el Toro and two bulk samples from Carbonera – 2005-2007

	Bulk Sample Description             & Source	Magnetic Concentrate       (Wt. %)	FeTOT           (%)	TiO2                 (%)	V2O5             (%)	Fe Recovery (%)
16.2 SGS Lakefield (2005)	16.3 Surface Raw and MC – Carbonera (Mineralogical Studies)	16.4 n/a	16.5 n/a	16.6 n/a	16.7 n/a	16.8 n/a
16.9 Midrex (2005)	16.10 Surface Raw Sample - Carbonera	16.11 11.43%	16.12 59.70%	16.13 2.02%	16.14 0.80%	16.15 69.22%
Solumet SGS (2006)	Surface Mag Concentrate - PET	n/a	66.20%	3.56%	0.56%	n/a
Eriez (2006)	Surface Raw Sample - PET	12.70%	53.36%	5.52%	0.43%	n/a
Midrex (2006)	Surface Raw Sample - PET	7.1%	59.11%	4.1%	0.45%	47.39%
Bateman (2006)	Borehole Raw Sample - PET	2.80%	55.86%	4.89%	0.47%	24.78%
Bateman    (2007)	Trench Raw Sample - PET Dry separation Wet Separation+milling	3.52% 3.35%	57.68% 60.32%	n/a	n/a	32.2% n/a

16.15.1

QSGS Lakefield (2005)

The primary objective was to ascertain the purity level of the magnetic concentrate, and determine whether grinding would be a viable exercise in order to upgrade the Fe content and decrease the amount of impurities.  Mineralogical investigations (optical microscopy, electron microprobe analysis and x-ray diffraction) were undertaken on the second-pass MC (#2 ROM) in order to ascertain the host of titanium in the ore, a penalty element or impurity when trying to produce a marketable iron concentrate (SGS Lakefield 2005a).

Mineralogical studies confirmed that only fine grinding to micron-scale would lower the TiO₂ concentration, which would be cost prohibitive from a production perspective.  The authors of the report recommend that cleaning of the Fe ore not by a physical process, but rather by a metallurgical process (eg. Fastmelt by Midrex) is a viable option.  The magnetic concentrate produced from these tests was forwarded to Midrex Technical Center for preliminary testing with Fastmelt.

16.15.2

Midrex Test (2005)

Based on positive results from preliminary testwork by Midrex (Ripke, 2005)), Cardero Resource Corp. supplied Midrex Technologies with a large bulk surface raw sand sample (1060 kg; August 2005) from the Carbonera dunes area for more comprehensive testing of magnetic concentrate processing (separation) and the feasibility of Fe concentrate upgrading utilizing the Fastmelt process.  Single pass magnetic separation of raw magnetite sands was deemed to be sufficient (12 in diameter, 5.5 in wide laboratory drum magnet separator at a feed rate of 41Kg/hour at 1.4 DC amps), and the bulk sample was processed in this manner.  The magnetic concentrate grade obtained was 11.43 Wt. %, with a total Fe content of 59.70% (of which 69.22% is recoverable), 2.02% TiO₂ and 0.80% V₂O₅ (Ripke 2005).

16.15.3

Solumet Upgrading Test (2006)

In order to independently assess the quality of the magnetic concentrate as well as any upgrading potential of the Fe ore, Worldlink Resources Canada appointed Solumet Metallurgical Solution Inc. (2006) to study a magnetic concentrate provided by Cardero Resource Corp.  The 2 tonne magnetic concentrate was produced from surface sands at Pampa El Toro using the dry magnetic processing facilities onsite in Lomas.  Magnetic concentrate grade recovery was not calculated by Cardero for this bulk sample.

Work conducted on behalf of Solumet by SGS Lakefield Research Ltd. concluded that upgrading by magnetic separation without grinding is feasible into a product with 75.5% Fe₂O₃ (52.8% Fe) and a recovery of 67% of the iron, and that further upgrading to 87.5% Fe₂O₃ (61.2% Fe) with an overall recovery of 58% of the iron is possible.  However, best concentrates achieved still contained between five and six percent TiO_2, which is detrimental to concentrate quality.  Mineralogy showed that the titanium is mainly contained in the form of ilmenite, intimately and finely exsolved in the crystal lattices of magnetite and hematite.  Liberation of such ilmenite for physical separation would require milling to a likely-uneconomic micron size, and separation of the titanium during melting remains probably the best solution.

Preliminary grinding and wet magnetic separation (Davis Tube analysis) tests were carried out on a sub-sample.  A process was then designed for the remainder of the bulk sample that involves grinding the concentrate to <180 microns, then passing the concentrate in a slurry over three low intensity magnetic separators (LIMS) to produce an upgraded product.

The result of this test is a concentrate containing 66.20% Fe^TOT (upgraded from 63.5%), 3.56% TiO₂, 0.56% V₂O₅ and 0.18% P₂O₅.  Grinding studies demonstrated that P₂O₅ concentration decreased with grinding, whereas TiO₂ and V₂O₅ concentrations remained roughly the same.  The findings corroborated previous mineralogical studies (SGS Lakefield, 2005a, 2005b; LeCouteur, 2005) which demonstrate that TiO₂ is encapsulated within the magnetite crystals as ilmenite lamellae, and, in the case of V₂O₅, within the magnetite crystal lattice itself.  Solumet states that, although the overall Fe content is adequately upgraded, the elevated concentrations of impurities are problematic.  Solumet stated that the product may be classified as a secondary quality Fe ore concentrate for direct shipping purposes.

16.15.4

Eriez Magnetic Separation Test (2006)

A 155kg sample of raw surface sands from Pampa El Toro was submitted to Eriez Magnetics Test and Research & Development, Pennsylvania (Eriez, 2006) in order to ascertain the optimum magnet and feeder set-up for processing the iron sands in addition to independently repeat the above Solumet test-work.  The results of this test were not definitive, as multiple tests were not carried out on several machines and compared.  Rather, the bulk sample was passed over a DFA-25 Drum at a setting of 10 tons per hour per foot (t/h/f) and drum speed of 500 feet per minute (f/m), and a second pass of the magnetic fraction at different settings (5t/h/f and 750 f/m).  This resulted in a magnetic concentrate grade of 2.67 Wt. %.  A sub-sample of the non-magnetic fraction from the first pass was passed over a SPRE 15-inch drum (settings not specified).  The result from this second test, back calculated, was 12.70 Wt. % magnetic concentrate.  Samples of each fraction were returned to Cardero, which were in turn analysed at Acme Laboratories in Vancouver.  The analytical results of the final magnetic concentrate were 53.36% Fe^TOT, 5.52% TiO₂ and 0.43% V₂O₅.

16.15.5

Midrex Test (2006)

A second bulk sample (435.6 kg) of Pampa El Toro raw sand was submitted to Midrex.  The mandate was to dry-magnetically-separate the raw sand in order to achieve a target minimum 62% Fe^TOT with favourable recoveries.  The processing equipment chosen for the test was an Eriez drum separator (type “A” – agitator type) with “Erium 25” axial magnetic fields using barium-ferrite ceramic permanent magnets.  This equipment was run on sub-samples at several different drum speed settings from which samples were split off and analysed to determine the optimum setting. 30 rpm produced the best grade of 60.1% recovered Fe.  A three-stage dry magnetic separation process was then designed.  The result of the test is 7.1 wt. % magnetic concentrate of the ROM that contains 59.1% Fe^TOT, 4.1% TiO₂ and 0.45% V₂O₅ with Fe recovery of 47.39%.  Recommendations in the report include the possibility of crushing the coarse size fraction of the magnetic concentrate in order to liberate non-magnetic material (silica and silicates adhering to magnetite) and potentially up-grade the product to the target grade of 62% Fe.

16.15.6

Bateman Engineering Test (2006)

After the phase II drilling program, a bulk sample (480 kg) of raw sand from a composite of ten drillholes (0-60 m depth) were submitted to Bateman Minerals and Metals (South Africa) for another magnetic separation process study (Rademeyer 2006).

Initial testing was performed on a small sub-sample of the drilling sands. 12.6 wt. % of the ROM reported to the sinks after heavy liquid separation (SG > 2.95).  The sinks were analyzed by XRF and contain 33.4% Fe₂O₃ and 3.38% TiO₂.  Particle size analysis was performed on both the ROM and sinks using wet screening techniques to reduce fines loss.  Results demonstrated that the prevailing particle sizes for the ROM and sinks are 180um and 125um respectively.  Dry magnetic separation on a sub-sample was performed (Carpo Laboratory Scale Induced Roll Lift Magnet) as well as optimization of the magnet strength to balance magnetic concentrate grade and Fe₂O₃ recovery.  The required strength was determined as 0.09 amps (81% Fe₂O₃, 5.44% TiO₂ with a magnetic concentrate grade of 2.9% of the ROM).

The bulk processing test was designed to assess the performance of an Eriez rare earth drum magnet (RED) in creating a high-iron magnetic concentrate.  After extensive testing, a process flow diagram was produced and recoveries and grades calculated.  The final recommendation for a plant is based on a three stage circuit: two passes through a RED magnet and a final electrostatic circuit upgrade to make a product grade of 79.8% Fe₂O₃ (55.86% Fe^TOT), 4.89% TiO₂ and 0.22% P₂O₅ from a magnetic concentrate grade of 2.75 wt. % of the ROM.

16.15.7

Bateman Engineering Test (2007)

A second bulk sample was sent to Bateman Engineering of raw sand from trenching at Pampa el Toro (622.8 kg).  The scope of this test was to determine the suitability of upgrading the sand with use of gravity spirals and follow-up wet magnetic separation, while at the same time compare these results to dry magnetic separation of a representative split of the ROM.

The dry magnetic separation process included 4 stages of magnetic separation.  The result for the best iron grade concentrate, albeit at lower recoveries, was after a simple two-pass process over an Eriez Model 380mm diameter BaFe drum magnet.

The spiral gravity separation tests were performed with a Multotec SC18 spiral at varying feed densities and rates.  This effectively removed much of the gangue material; however a further upgrading by wet magnetic separation was necessary due to the large amount of heavy non-magnetic minerals.  The wet magnetic separations of the concentrate were done in a slurry of water.  The resulting concentrate contained 50% Fe, and a final milling stage was tested to further upgrade the material.  With the possibility of further refining the milling stage parameters (as this step was not part of the original work proposal), the final concentrate was upgraded to 60.32% Fe with a mass recovery of 3.35%.

16.16

Mineral Processing and Metallurgical Testing 2008

16.16.1

Mineral Processing

The major equipment supply for a dry magnetic separation pilot plant was provided by Eriez Manufacturing Co. with input from Cardero Resource Corp., based on the small-scale laboratory magnetic separation of Pampa El Toro iron sand feed conducted during 2005 and 2006 as reported previously.  The equipment packages were then fabricated and delivered to the Pampa El Toro pilot plant site in late 2007.  The plant site is located at the base of the Pampa El Toro dune near the village of Acari.

Pilot plant operations were carried out under the supervision of the Cardero technical team (Hoffman and Ripke), and consultants from Holland & Holland and BBA Engineering.

The pilot plant facility was commissioned in March 2008 during Phase I of the pilot plant test.  Two additional periods of pilot plant operation occurred in 2008: Phase II (May) and Phase III (June).  A final total of 56.8 tonnes of Pampa El Toro magnetic concentrate was produced (Table 15.2).

Figure 16.1: Illustrated photograph of the Eriez magnetic separation pilot plant, located in Acari.

The dry magnetic separator pilot plant consists of three core components: a primary separator, a scavenger separator and a final cleaner separator, along with various associated hoppers and conveyor systems (Figure 16.2).  A flow sheet is presented in Figure 16.2.

Figure 16.2: Flow sheet illustrating the magnetic separation process of the pilot plant.

Paired samples of concentrate and tails from each magnet, as well as the initial feed samples were collected on an hourly basis during pilot plant operation.  Initial analyses were conducted onsite with a portable XRF analyzer (Innov-X XRF Alpha 6000 Analyzer) to aid the optimization and control of the plant.  As a check on the XRF results, representative splits of a subset of these samples were sent to Acme and ALS-Chemex laboratories for analysis and the remainder of the samples are being stored at the secured site.  All Fe grades reported in Table 16.2

Summary table of amount of sand processed, magnetic concentrate produced, weight recoveries and concentrate grades from each of the three pilot plant tests. (std = standard deviation) are analytical laboratory results.

During commissioning of the pilot plant (Phase I), approximately 83 tonnes of the feed sand was processed, producing 4.8 tonnes of iron concentrate at 47.9±2.4% Fe (Table 16.2

Summary table of amount of sand processed, magnetic concentrate produced, weight recoveries and concentrate grades from each of the three pilot plant tests. (std = standard deviation)).  Feed rates, magnetic separator drum speeds, splitter positions, and the overall flow sheet were all variables tested during commissioning (Hoffman, 2008a; Holland, 2008).

During Phase II, the pilot plant throughput was improved from 2.8 to 27 tonnes per hour (Hoffman, 2008b).  The scavenger magnetic separator recovered only a low grade iron material as its C2 concentrate, so it was discarded rather than send it to the finisher magnetic separator.  Consequently the scavenger magnetic separator was removed from the flowsheet.  The average feed grade for the head sand of the ROM was determined to be 7.3±1.7% Fe, however additional testing is recommended.  A total of 46 tonnes of magnetic concentrate at a grade of 55.5±2.8% Fe was produced from 613 tonnes of head sand resulting in a weight recovery of 7.5% MC.  The concentrate was produced during nearly continuous operation for eight hours per day over a 10 day period at steady-state parameters (Hoffman, 2008b).  The calculated Fe recovery for the plant is 63.66%, based on data gathered from this period of steady-state operation (Hoffman, 2008b).

During Phase III, the pilot plant was modified to increase the feed throughput to 30 tonnes per hour.  The maximum plant throughput was investigated by batch processing.  A throughput of 30 tonnes per hour was achieved with essentially no loss in recovery on the primary magnet.  The concentrate weight recovery is 7.3 weight per cent as an average of nine separate batch tests, which compares to the 7.5% determined in Campaign II.

Table 16.2: Summary table of amount of sand processed, magnetic concentrate produced, weight recoveries and concentrate grades from each of the three pilot plant tests. (std = standard deviation)

Pilot Plant Phase #	Raw sand feed type	Tons of raw sand processed	Tonnes of magnetic concentrate produced	Magnetic concentrate Weight recovery, %	Concentrate Grade, % Fe (±1 std)
I	PET	83	4.8	5.8	47.9±2.4
II	PET	613	46.0	7.5	55.5±2.8
III	PET	82	6.0	7.3	57.1
	16.17 Carbonera	16.18 200	16.19 19.0	16.20 10.6	16.21 59.9±0.2
Total	PET	778	56.8	7.3

Preliminary bench-scale test work, completed by RDI Consultants, Denver, and followed up by Natural Resources Research Institute (NRRI), indicates that the iron concentrate can be further upgraded to 62.8% iron by screening out a portion of the concentrate that has a high relative concentration of silica (Table 16.3).

One tonne of iron ore concentrate has been shipped to the Natural Resources Research Institute (NRRI), Minnesota and test work is underway to further concentrate it through screening and other mineral separation techniques, targeting the production of a concentrate grade of 64% iron, which would be suitable for blast furnace sinter feed production.  It is noteworthy that the total iron content of the concentrate with the +80 mesh fraction removed calculates to 64.7%.  This top fraction can either be rejected or ground slightly to recover this iron at a higher grade by recycling it as a portion of the feed to the cleaner magnetic separation.

Table 16.3: NRRI in-house laboratory results from the screening test on Pampa el Toro magnetic concentrate.

Screen Size, mesh	Wt (%) retained	FeTot	Fe++	CaO	MgO	Al2O3	SiO2	TiO2
65	6.01	21.31	5.31	5.04	3.09	14.15	48.31	2.91
80	5.32	46.94	12.35	2.13	2.1	3.58	20.97	5.82
100	11.42	60.79	16.19	2.13	2.1	3.58	8.02	5.82
150	33.08	65.31	17.77	0.78	1.09	1.51	3.92	6.11
200	32.87	66.98	17.34	0.51	0.88	1.34	2.83	6.27
270	8.57	62.53	14.62	0.79	1.22	2.34	6.47	6.12
400	1.91	51.53	10.58	1.64	1.97	4.38	14.3	6.68
-400	0.81	51.53	10.58	1.64	1.97	4.38	14.3	6.68
Head Calculated		61.11	15.94	1.20	1.35	2.71	8.11	5.94
Head Analysis		60.04	15.29	1.21	1.36	2.63	8.18	5.69

16.21.1

Metallurgical Testing

The Phase I briquetting work and the Phase II bench-scale pig iron production program has been successfully completed by NETL. The Phase III pilot-scale pig iron production program has been completed by NETLand the analytical results are currently being determined.

In the Phase I briquetting and Phase II bench-scale pig iron production work, a portion of the Pampa El Toro concentrate was agglomerated, processed into liquid hot metal (molten iron) and cast into pigs in a total of ten bench-scale electric arc-based smelting furnace tests. A total of 128 kilograms of pig iron were produced. The test work was conducted in cooperation with the U. S. Department of Energy (DOE)'s National Energy Technology Lab (NETL) located in Albany, Oregon. The weight recovery of pig iron was very good at up to 90%.

The certified chemical analysis of the sulphur and carbon content of the pig iron was determined by Howmet Research Center (Whitehall, Michigan, USA). The chemical analysis of the remaining elements of interest was determined at NETL by XRF. NETL’s chemistry lab is not certified, so an alternate certified lab is being identified to verify these analytical results. The optimum bench scale test, test run number 10 of the 10 bench-scale tests was selected for subsequent larger pilot-scale testing. For test #10, the furnace was fed with 50.0 kg of iron ore and additional weight of reductant, slag, and binder and yielded 26.5 kg of pig iron and 11.7 kg of slag.  The following table shows the chemical analyses for the iron ore concentrate that was processed into pig iron and slag for test # 10 (Table 15.4). The results were positive and either conform to or exceed industry specifications.

Table 16.4: Chemical analyses for bench test #10: pig iron, iron ore and slag.

%	Fe	C	S	Mn	Si	Al	V	Ti	Other	Total
Pig Iron	96.90	2.61	0.085	0.050	0.070	0.020	0.20	0.018	0.047	100

%	Fe2O3	MnO	SiO2	V	TiO2	Al2O3	CaO	MgO	Other	Total
PET iron ore	77.55	0.35	10.6	0.257	4.86	3.17	1.43	1.54	0.293	100

%	Fe	C	S	Mn	SiO2	V	TiO2	Al2O3	CaO	MgO	Na2O	K2O
Slag	1.38	0.03	0.11	0.991	39.0	0.403	16.00	18.85	14.00	5.14	1.56	1.115

Figure 16.3  Photograph of the products of the smelting tests performed on the Pampa el Toro concentrate.

The pig iron product from test #10 has undergone chemical analysis and quality was determined to be good. Additional bench-scale smelting tests may be conducted following receipt of results. Knowledge gained from the bench scale tests have been incorporated in the large scale pilot smelting test.

The Phase III pilot-scale smelting tests were completed in July 2009 utilizing the optimum briquette formulation and operating control parameters as determined from the bench-scale smelting tests discussed above. The pilot-scale smelting test successfully demonstrated consistent smelter operation, uniform hot metal quality and slag quality over an extended period of operation time. The preliminary chemical analyses are currently being confirmed by an independent certified lab.

17

Mineral Resource and Mineral Reserve Estimates

Cardero supplied SRK with a set of data from their geological investigations, exploration program and surface surveys including the following:

• Reports on the geology and the exploration programs;
• Drillhole collar positions and assay results;
• Bulk Density and Specific Gravity testing methodologies and results; and
• Contours of the surveyed topography elevation.

Based primarily on this data SRK generated a Mineral Resource estimate for a portion of the Pampa el Toro dune field. The data analysis and Mineral Resource estimation is described in this section.

17.1

Data Statistics

SRK were supplied with the collar coordinates and assay results for 112 drillholes. The assays results are from 5m composite samples and extend to 30m depth for most holes (6 holes have assays extending to 60m). The majority of the holes were drilled to depth of 60m (103 holes), one hole (PET-12) stopped at 38m on bedrock, and the remainder of the holes were drilled to 100m or greater depth. Samples were collected from the drill rigs at 1m intervals and composited in the field to create the 5m composite samples. The details of the compositing are contained in Section 12. At present, only samples from the top 30m have been submitted for analysis, as it was estimated that the resource above this level would contain sufficient material to complete the necessary feasibility studies.

A total of 718 assay results were supplied to SRK with analyses for a wide range of elements, including the major element oxides by ICP-OES and minor elements by ICP-MS as discussed in Section 13. SRK have focused its analysis on the three elements of economic interest (Fe, Ti and V), but has also considered a number of other variables that may impact the project (such as Al₂O₃, SiO2, MgO, MnO, K₂O, Na₂O, and P).

SRK conducted an analysis of the univariate statistics of the selected set of variables, as well as the correlation between the variables. The univariate statistics are presented in Table 17.1 and a correlation matrix is presented in Table 17.2. All variables have low a Coefficient of Variation (“CoV”) reflecting the relatively consistent distribution of the elements in the deposit, as well as the approximately normal distributions of all of the variables. MnO, MgO, Na₂O and Al₂O₃ have very low CoV and this reflects the limited range of values in the assay results.

Histograms of the selected variables are presented in Figure 17.1, and Scatter Plots of %Fe₂O₃, %TiO₂, Vppm, and %SiO₂ are presented in Figure 17.2.

Table 17.1: Univariate statistics of selected variables from the 5m composites.

Variable	Count	Minimum	Maximum	Mean	Standard Deviation	Variance	Coefficient of Variation
Fe2O3 %	718	3.36	13.6	6.63	1.45	2.11	0.22
TiO2 %	718	0.39	1.48	0.72	0.15	0.02	0.21
V ppm	718	4.0	360	170.1	43.8	1919.5	0.26
MnO %	718	0.1	0.2	0.1	0.02	0.00	0.17
MgO %	718	1.2	2.6	1.8	0.23	0.05	0.13
P2O5 %	718	0.09	0.19	0.13	0.02	0.00	0.11
K2O %	718	1.6	2.24	1.99	0.09	0.01	0.05
Na2O %	718	3.06	4.17	3.46	0.18	0.03	0.05
Al2O3 %	718	14	16.66	15.26	0.44	0.19	0.03
SiO2 %	718	56.78	68.17	63.92	1.55	2.40	0.02

The correlation coefficients in Table 17.2 have been colour coded with the positive correlations having hot colours, and the negative correlations having cooler green colours. The correlation between %Fe₂O₃, %TiO₂ and Vppm is very strong, suggesting that either the metals are contained within the same minerals, or that the deposition processes have acted on the host mineral in a very similar manner. MgO and MnO also show good correlations with %Fe₂O₃, %TiO2 and Vppm suggesting that these metals are likely to be contained within the magnetite, titano-magnetite and ilmenite crystals. %P₂O₅ shows a reasonably good correlation with the 5 aforementioned variables, although the P is likely to be contained within Apatite grains. Al₂0₃, SiO₂, K₂O, and NaO₂ all show relatively strong negative correlations with %Fe₂O₃, %TiO₂, Vppm, %MgO, and %MnO, and relatively weaker negative correlations with P₂O₅. SiO₂ in particular has a strong negative correlation with %Fe₂O₃, %TiO₂, Vppm, %MgO, and %MnO, suggesting that Quartz is the main gangue mineral.

Table 17.2: Correlation Matrix of selected variables from the 5m composites.

Figure 17.1 on page 59present histograms of the analyzed variables and Figure 17.2 presents scatter plots of  Fe₂O₃, TiO₂, V and SiO₂ indicating the strong correlations that exist between the variables. The histogram bars in red in Figure 17.1 are the samples that were masked during the calculation of the experimental semi-variograms discussed in section 17.2.

Figure 17.1: Histograms of grade variable for selected variables from the 5m composites.

Figure 17.2: Scatter plots of Fe₂O₃, TiO₂, V and SiO₂ from the 5m composites indicating the strong correlations that exist in the data.

17.2

Semi Variogram analysis

The 5m composites from the drillholes were used to generate experimental semi-variograms in a plane that is approximately parallel to the average topography. The variations in the topography and the large distances between samples made the selection of an appropriate plane difficult, however after some trial and error SRK settled on a plane with a dip of 3.3º to the south. SRK tested the data for indications of anisotropy within the plane selected, by calculating semi-variograms in 10º sectors, and assessing the variations in continuity. The direction of best continuity is east-west; however the variogram in a north-south direction shows a very similar structure, with some evidence of a trend. Additionally, the semi-variogram in the east-west direction did not reach the population variance, until after a stable ranged had been reached, and showed no interpretable structure at that range, as can be seen in Figure 17.3.

Figure 17.3: Directional semi-variogram for Fe₂O₃ in the plane approximately parallel to the topography

After the masking of four holes in the south of the deposit (plotted in Figure 17.4), which were mostly drilled at significantly wider distances than the 500m grid, the trend structure disappeared from the experimental semi-variogram, and more robust structures could be resolved and modeled. One of the masked holes was a twinned hole with unusually high grades that contributed to high variances in the experimental semi-variogram. It is likely that at the distances between the widely spaced holes in the south and the main concentration of data, as well as changes in the topography and average dip of the surface mean that the sample pairs which generate the semi variogram at longer distances are from different layers, which have less correlation than samples from the same layers. The assays from the four holes (PET011, PET012, PET010 and PET008) that were masked from the dataset used for generation of the semi-variogram, were included in the dataset used for Mineral Resource estimation. The hole collars with red symbols in Figure 17.4 are the masked holes; the crosses symbolizing the collar locations are shown proportional to the drilled depth of the hole.

Figure 17.4: Plan view of drillhole collars displaying masked drillholes in red

The structures of the directional semi-variograms in Figure 17.3 before the 1500m distance on the X Axis are very similar, and SRK elected therefore to model only an omni-directional semi-variogram in the selected plane as this yielded more robust structures, and the scale of possible anisotropy is not considered significant. SRK modeled a shorter range semi-variogram in the direction perpendicular to the sub-horiontal plane which approximates a down hole semi-variogram. In the plane of the omni-directional semi-variogram, a slicing height of 7.5m was used. Because of the good correlation between the major elements (albeit negative correlations such as that between Fe₂O₃ and SiO₂) most of the semi-variograms show very similar ranges and structures. Examples of the semi-variograms modeled for Fe₂O₃, TiO₂, V and SiO₂ are presented in Figure 17.5.

Figure 17.5: Omni-directional semi-variograms in the plane parallel to the topography surface, for Fe₃O₂, TiO₂, V and SiO₂

In all cases two structured spherical models were fitted to the experimental data in the plane parallel to the topography, and in the direction perpendicular to the omni-directional plane. Table 17.3 presents the modeled semi-variogram parameters for all modeled variables. All variables except Na₂O, Vanadium and Al₂O₃ have models with a second range of over 2000m indicating long range continuity in the mineralization. The total sills of all the models except Na₂O also approximate the population variance. In all cases there is a relatively high nugget effect (modeled from the down-hole semi-variogram) of between 50 and 70% of the total sill.

Table 17.3: Modeled semi-variogram parameters for all variables estimated

First Structure Second Structure

	C0	C1	Range 1	Range 2	Range 3	C2	Range 1	Range 2	Range 3
Fe2O3 %	1.342	0.142	292	292	12	0.51	2468	2468	18
TiO2 %	0.013	0.0016	438	438	12	0.006	2845	2845	22
V ppm	1193.4	187.1	341	341	10	422	1812	1812	17
SiO2 %	1.419	0.14	1000	1000	15	0.675	2331	2331	25
P2O5 %	0.000	0.00004	500	500	30	0.00006	2200	2200	45
Na2O %	0.011	0.0024	150	150	10	0.0089	840	840	23
K2O %	0.005	0.0007	147	147	15	0.0016	1775	1775	28
MgO %	0.023	0.0041	590	590	20	0.0218	2587	2587	27
MnO %	0.0002	0.00001	773	773	18	0.0001	2331	2331	28
Al2O3 %	0.080	0.0305	292	292	10	0.0450	961	961	18

To test the robustness of the semi-variogram model and the search parameters SRK conducted a Cross Validation process for the three variables of economic interest. The cross validation procedure considers each composite in the database sequentially, and temporarily removes the composite from the database, before Kriging the data point using the defined semi-variogram and search parameters. The estimated value is then compared with the true value. Outputs from the Cross Validation include the mean error (Z*-Z) and the mean standardised error ( ), which measure the degree of unbiasedness. A graphical representation of the results is presented in Figure 17.6, Figure 17.7, and Figure 17.8 for Fe₂O₃, TiO₂, and V respectively.

The mean (Z*-Z/S*) values for Fe₂O₃, TiO₂, and V are -0.004, -0.006, and -0.003 respectively. This indicates that there no material bias in the estimates, and the majority of the composites estimated values compare favourably with the real values, with a small number of composites (between 8 and 11 composites) falling outside the selected threshold of 2.5 (interval [-2.5 ; 2.5]) which defines outliers as being outside the 99% confidence limit of a normal distribution. This tendency to underestimate some of the highest grade values, however is considered to be a normal consequence of the smoothing effect that results from the Kriging process.  The Cross Validation tests indicate that with the selected search criterion (discussed in Section 17.5) the estimates can be considered robust.

Figure 17.6: Cross Validation outputs for Fe₂O₃ %, from the 5m composites.

Figure 17.7: Cross Validation outputs for TiO₂ %, from the 5m composites.

Figure 17.8: Cross Validation outputs for Vanadium ppm, from the 5m composites.

17.3

Density Determination

Samples from nine drillholes were selected for bulk density determinations using a graduated cylinder from which the samples volume and weight were measured. Two bulk density determinations were done for each 5m composite sample from the 9 drill holes (resulting in two sets of 54 determinations each), one on untreated sand, and another on compacted sand that had been vibrated for 15 minutes on a vibrating plate. Histograms for the compacted and un-compacted samples are presented in Figure 17.9. The compacted bulk densities appear to have a bimodal distribution, however; there is no apparent correlation between major element concentration, or depth, and the range of values is tightly constrained, having a coefficient of variation of 0.04. This may be a result of the small number of samples rather than two populations of data. SRK consider the average value of 1,81 t/m³ to be an acceptable estimate of the average bulk density for the deposit, and this has been used to calculate the tonnages in the Mineral Resource estimate.

Figure 17.9: Histogram of compacted (left) and un-compacted (right) Bulk Density determination results

17.4

Wireframe modeling

Cardero supplied SRK with a set of contour lines that defined the topographic surface over the major portion of the deposit. The topographic contours were used as the basis for generating the topography wireframe, which is used to vertically constrain the Mineral Resources. A comparison of the contour elevations and the collar elevations of the drillholes reveals that there are some significant differences between some drillhole data and the surface defined by the contours. In consultation with Cardero, SRK elected to project the collar positions onto the wireframe surface. Where there were no contours, SRK relied on the collar elevations of the drillholes to model the surface topography. The topography surface was therefore created in the following sequence:

• A wireframe surface was created based on the contour stings only;
• The drillhole collars that overlapped the wireframe were vertically projected onto the wireframe surface; and
• A second surface was created, using the contours and the borehole collar positions. Collar elevations within the area covered by the contours are consistent with the surface elevation, while the surface is defined by only the collar elevations outside of the area covered by contours.

As discussed in Section 17.1 the majority of holes were drilled to at least 60m or deeper, although only samples to a depth of 30m were typically assayed. The mineralization is assumed to be continuous below 30m depth, which is supported by the results from the six holes that have assays to 60m. In these 6 holes there is no indication of significant changes in the grades of the primary variables (Fe₂O₃, TiO₂ and V). The topography surface was copied 30m vertically downwards and used to constrain the depth extent of the Mineral Resource.

17.5

Resource estimation

A three dimensional block model was created to model the volumes of mineralized material, and to contain the grade estimates. The rotation that was applied to the semi-variograms (see section 17.2) was also applied to the block model such that the XY plane was approximately parallel to the topographic surface. The blocks were defined as 250m in the X and Y plane, as this resulted in one drillhole in the approximate center of each block in the densely drilled area, and one drillhole in every alternate block in the more sparsely drilled areas. The blocks are generated with a 5m Z axis dimension, which matches the composite lengths, and would typically result in one sample per block where a drillhole exists in a block column.

The block model was split to better model the volumes defined by the topography surface and the depth boundary of the Mineral Resource. The blocks were sub-celled to create the minimum number of blocks (i.e. where an entire parent cell did not intersect a wireframe, it was not split, however where it was split, the largest blocks possible were created). The minimum block size in X or Y was set at 6.25m (or 1/40^th of the parent block size), whereas the in the Z axis blocks were split to precisely match where the wireframe crossed the block axis. In the estimation however, only the 250m parent blocks were estimated (i.e. all sub-blocks within an original parent block would receive the same grade estimate).

SRK conducted a kriging search neighborhood analysis exercise for Fe₂O₃ to determine the optimum number of samples to use in the estimates. A single block in a densely sampled area is repeatedly kriged using the same parameters, and altering only the maximum number of samples in each cycle. Various indicators of the quality of the kriged estimates, such as the Slope of Regression (“SR”) and Kriging Efficiency (“KE”), are compared. The results of the tests are illustrated in Figure 17.10.

Figure 17.10: Kriging search neighbourhood analysis results

Increasing the number of samples used in the estimate generally increases the quality of the estimate. In Figure 17.10 the KE and SR consistently increase as more samples are included in the kriging neighborhood. The estimation variance also decreases as more samples are included. Using a very large number of samples in the estimate can result in an overly smoothed estimate, where none of the local variability in the deposit is retained. SRK selected 12 as the maximum number of samples to use in the search neighborhood as this resulted (for the block being estimated) in a SR of over 0.8 and a Kriging efficiency of over 0.6. Additionally, for a typical block, with a drillhole in the center of the block, using 12 samples, the samples selected will typically have the configuration schematically illustrated in Figure 17.11.

The block to be estimated is the central red block, while the surrounding blocks are the adjacent blocks in the same plane as the block to be estimated. The 5m samples from the drillholes that pass through each of the blocks are represented as yellow cylinders, with the central sample in each drillhole in the same plane as the block being estimated. Samples marked with an X are the samples that are be selected by the search neighborhood definition. The sample at the center of the block to be estimated is selected, as well as the other 8 samples in the plane of the block to be estimated. Additionally one other sample from the drillhole passing through the block to be estimated (above or below the block) and two other samples from the central cross of drillholes (depending on which drillhole is closer to the center of the block being estimated) are selected.

Figure 17.11: Schematic illustration of the samples selected in a typical search neighborhood

SRK used Ordinary Kriging to populate the blocks with grade estimates for a range of variables that could be economically important, or could possibly affect the mineral processing. The variables estimated are Fe₂O₃ %, TiO₂ %, V ppm, SiO₂ %, P₂O₅ %, Na₂O %, K₂O %, MgO %, MnO %, and Al₂O₃ %. Block discretisation of 10 by 10 by 5 was used in the estimation. Any blocks not estimated for Fe₂O₃ %, TiO₂ %, V ppm were excluded from the Mineral Resources, and only Fe₂O₃ %, TiO₂ %, V ppm are reported as Mineral Resources. Average estimated grades of the other estimated variables are reported separately from the Mineral Resources. Only the parent 250m by 250m by 5m blocks were estimated, and all sub blocks within the parent blocks have the same grade value.

17.6

Classification and Mineral Resource Reporting

The Mineral Resources estimated by SRK for the Pampa el Toro are reported in Table 17.4. The Mineral Resources are classified in accordance with the Canadian Institute of Mining, Metallurgy and Petroleum (CIM) Standards on Mineral Resources and Reserves – Definitions and Guidelines. SRK used a scorecard approach in the classification:

• Geological logging and sampling quality: SRK conducted site visits in 2006 and 2008 to assess the quality of geological logging and sampling and were satisfied that due care was being taken in the logging, handling and sampling of the percussion drilling samples.
• Data quality: Cardero have implemented industry standard procedure for data collection and storage during the exploration program.  The quality control program implemented by Cardero is sufficient to monitor the accuracy and precision of the analytical results. The analysis of the quality control sample results indicates sufficient accuracy and precision on the analytical results.
• Geological interpretation: The geological model proposed by Cardero for the formation of the dune sands is supported by the exploration results, as well as by the results seen in the Mineral Resource estimation.
• Geological modelling: The geological model generated by SRK honours the drilling information supplied by Cardero. There is little extrapolation of the model beyond the 500m spaced drilling.
• Grade estimation: The experimental semi-variograms calculated by SRK show good continuity, and ranges significantly longer than the average drillhole spacing, and SRK consider the grade estimates to be robust.

Table 17.4: Mineral Resources for the Pampa el Toro project as at 21 July 2009

Category	Volume M3 '000	kT	Fe2O3 %	TiO2 %	V ppm
Indicated	133,608	241,831	6.67	0.72	172
Inferred	348,190	630,224	6.47	0.70	166

1 No grade cut-off has been applied to the material in the Mineral Resource, as the grades are relatively homogenous, would likely be able to be blended, and are projected to be economically viable, based on the assumptions made by Cardero. A Mineral Resource is not a Mineral Reserve and does not have demonstrated economical viability.

SRK consider that all aspects of the exploration program have produced a high quality dataset with good external control on the analytical results. The quality of the estimates, measured by the Slope of Regression of the actual value knowing the estimated value of the specified variable, and the Kriging efficiency. The portion of the deposit drilled on a 250m grid has slopes of regression that generally exceed 0.7, and Kriging efficiencies generally in excess of 40%. SRK has classified the one surrounding the 250m drilling as Indicated Mineral Resources.

The portion of the deposit surrounding this, drilled on approximately 500m centers, and extending approximately 500m beyond the area drilled on 500 centers is classified as an Inferred Mineral Resource. Portions of the deposit that were estimated further than 500m from the 500m spaced drilling were also excluded from the mineral Resources until further confirmatory work is completed to confirm the grades of the material.

The areas defined as Indicated and Inferred Resources are illustrated in Figure 17.12

Figure 17.12: Classified Mineral Resources with drillhole collar locations

18

Other Relevant Data and Information

No other relevant information has been identified

19

Interpretation and Conclusions

Exploration conducted to date at the Iron Sands Project has resulted in the discovery of a large, low-grade magnetite sand deposit at Pampa El Toro which could be amenable to modern bulk mining technologies.  Surface sampling and subsequent percussion drillhole testing indicates that the overall magnetite content is relatively homogeneous, particularly in the uppermost 30 meters.  Sample density achieved in the investigations is considered to be more than adequate to support these assertions.

Based on available information, it would appear that a certain amount of milling and grinding of the sand has occurred during percussion drilling, resulting in samples which may result an under-estimation of magnetic mineral content and Fe grades due to possible loss of magnetite to fines in addition to contamination of MC by fine non-magnetic fractions.  This is not expected to affect the in situ Fe grade which the current Mineral Resource estimate is based.  

Mineral processing/recovery tests conducted by five separate, independent laboratories provide encouraging results clearly indicating the feasibility of magnetic separation and upgrading of raw sand material to produce a final iron concentrate of 53.4% to 66.2% Fe.  The tests have investigated various upgrading techniques with the objective of reducing the Si content of the concentrate.  These included dry magnetic separation using a variety of drum magnet specifications, gravity spirals, wet magnetic separation techniques and grinding of the concentrate.  Whereas the tests that included wet separation techniques and grinding of the concentrate were able to upgrade the material, their economic viability outweighs their effectiveness. Dry magnetic separation techniques were determined to be the most effective and viable.

The Pilot Plant dry magnetic separation tests recovered up to 7.5 Wt. % MC at a grade of 55.5% Fe from the Pampa el Toro bulk sample.  Screening of the concentrate is currently being investigated as an additional upgrading step, as preliminary testing has indicated that the grade may be improved to 62.8% Fe.  As this would be a simple and economic upgrading step, further testing is strongly recommended.

A robust Mineral Resource estimate has been generated based on a sound geological model and exploration database. The Mineral Resource is classified in terms of the CIM definitions on Standards on Mineral Resources and Reserves – Definitions and Guidelines. An Indicated Mineral Resource of 133,608 kT with a Fe₂O₃ grade of 6.67% has been defined, along with an Inferred Resource of 348,190 kT with a Fe₂O₃ grade of 6.47%.

Laboratory scale metallurgical testing indicates that it is feasible at the bench-scale to produce a high quality, saleable, iron product.  Pilot plant scale metallurgical testwork is required to determine if this is possible at an industrial scale.

20

Recommendations

SRK recommend that the metallurgical test work continue to improve the confidence in the ability to upgrade the raw sand material into a saleable product, and that Cardero proceed with the preliminary economic assessment of the project.

21

References

Belik, G. (2005)  Summary of Peru Magnetite Sands Project for June 2005.  Internal report to Cardero Resource Corp., July 2005.

Cook, S. (2006): Report on 2005-2006 Exploration work on the Peruvian Iron Sands Project: Pampa El Toro and Carbonera Dune Fields, Southern Peru.  Report for Cardero Resource Corp., 14 June 2006. 256 pp.

Cook, S. (2006): Update on Peru Iron Sands Project (Draft).  Short report. 4 pp.

Eriez Magnetic Research and Development (2006).  Test summary: Magnetic concentrate of Black Sands. 14 September 2006. 3 pp.

Gay, S.P. Jr. (2005).  Blowing Sand and Surface Winds in the Pisco to Chala Area, Southern Peru.  Journal of Arid Environments, Volume 61, pages 101-117.

Hoffman, G. (2008a) Final PET Pilot Plant Test Report. Internal report to Cardero Resource Corp., April 2008. 9 pp.

Hoffman, G. (2008b): Campaign 2 PET Pilot Plant Test Report and RDI C3. Internal report to Cardero Resource Corp., July 2008 5 pp.

Holland, L. (2008) Preliminary report on the Pilot Plant Operation of Pampa el Toro. Report to Cardero Resource Corp, March 2008. 7 pp.

Hawkes, N., Clark, A.H. and Moody, T.C. (2002).  Marcona and Pampa de Pongo: Giant Mesozoic Fe-(Cu, Au) Deposits in the Peruvian Coastal Belt.  In:  Porter, T.M. (Ed.), Hydrothermal Iron Oxide Copper-Gold and Related Deposits:  A Global Perspective, Volume 2. PGC Publishing, Adelaide, pages 115-130.

Le Couteur, P.C. (2005).  Comments on Extracting Magnetite from Iron Sand Samples, Nazca area, Peru.  Micron Geological Ltd.  Preliminary report to Cardero Resource Corp., December 25, 2005. 14 pp.

Le Couteur, P.C. (2006).  Report on a visit to Cardero´s Iron Sands Project Area, Nasca area, Peru.  Micron Geological Ltd.  Preliminary report to Cardero Resource Corp., January 22, 2006. 11 pp.

O’Connor Parsons, T. (2006).  Pampa El Toro Fe Sands Project – Update.  Cardero inter-office correspondence, August, 2006. 4 pp.

Rademeyer, L.P. (2006).  Pampa El Toro Titano-Magnetite Deposit, Peru (Dry Magnetic Test Work).  Report on behalf of Bateman Minerals and Metals (Pty) Ltd. for Cardero Resource Corp. 29 November, 2006. 23 pp.

Rademeyer, L.P. (2007). Pampa El Toro Titano-Magnetite Deposit, Peru (Dry and Wet Magnetic Test Work).  Report on behalf of Bateman Minerals and Metals (Pty) Ltd. for Cardero Resource Corp. 13 September, 2007. 19 pp.

Ripke, J. (2005).  Final Report – Magnetic Separation of Koripampa Sand and Fastmelt Test Series II Results.  Midrex Technical R & D Centre.  Report to Cardero Resource Corp., November 30, 2005. 12 pp.

SGS Lakefield Research Limited (2005a).  A Mineralogical Examination of  #2 ROM Peruvian Sands.  SGS Lakefield Research Limited.  Report to Cardero Resource Corp., April 11, 2005. 18 pp.

SGS Lakefield Research Limited (2005b).  An investigation into “The Recovery of Magnetite from a Mineral Sand Deposit in Peru”.  LR 10987-001-Report Nº 1. 15 April, 2005. 10 pp.

Solumet Metallurgical Solution Inc. (2006).  Technical evaluation of magnetite concentrate from Cardero´s “Iron Sands” in Peru.  A report prepared on behalf of Worldlink Resources Canada, in cooperation with Cardero Resource Corp.  21 September, 2006. 25 pp.

Torres, S. (2006a).  Informe de la Segunda Fase del Programa de Perforación en Pampa El Toro.  Mayo – Julio del 2006.  Internal report to Cardero Resource Corp. October 2006. 78 pp.

Torres, S. (2006b).  Trench Sampling Method.  Inter-office memorandum to TOCM. 12 September, 2006. 4 pp.

Van der Merwe, A. (2006a).  Report on a Site Visit to the Pampa El Toro and Carbonera Iron Sands Deposits, Ica Region, Peru.  SRL Consulting, Report 360457 prepared for Cardero Resource Corp.  April, 2006. 23 pp (Draft

Van der Merwe, A. (2006b).  Peru site visit June 2006. Memorandum from SRK Consulting to Mark Cruise at Cardero Resource Corp.  29 June, 2006. 8 pp

22

Date and Signatures

[Item 24 of FORM 43-101F1]

Effective Date of Technical Report: 10 September 2009.

Date of Signature: 10 September 2009.

Signed by:

_______________________________________

Mark Wanless (Pr.Sci.Nat.)

Principal Geologist

Mark D. Wanless

SRK Consulting (South Africa) (Pty) Ltd

265 Oxford Road

Illovo

2196

South Africa

Telephone: +27 11 441 1111

Fax: +27 11 880 8086

Email: mwanless@srk.co.za

CERTIFICATE of AUTHOR

I, Mark David Wanless, Pr. Sci. Nat. do hereby certify that:

1. I am a Principal Geologist of:

SRK Consulting (South Africa) (Pty) Ltd

265 Oxford Road

Illovo

2196

South Africa

2. I graduated with a B.Sc. degree in Geology, and Environmental and Geographical Science from the University of Cape Town in 1994. In addition, I have obtained a Bachelor of Science with Honours (Geology and Geochemistry) degree (1995) from the University of Cape Town.

3. I am a Professional Earth Scientist registered with the statutory body South African Council for Natural Scientific Professions.

4. I have worked as a geologist for a total of 14 years since my graduation from university.

5. 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 fulfil the requirements to be a “qualified person” for the purposes of NI 43-101.

6. I have overall responsibility for the technical report titled “Pampa el Toro Mineral Resource Technical Report”, and specifically sections 14.2 and 17 to 22 and dated 10 September 2009 (the “Technical Report”) relating to the Pampa el Toro project. I visited the Pampa el Toro project from the 21^st till the 23^rd of February 2008 for 3 days.

7. I have had no prior involvement with the property that is the subject of the Technical Report.

8. At the date hereof, to the best of my knowledge, information and belief, the Technical Report contains all scientific and technical information that is required to be disclosed to make the Technical Report not misleading.

9. I am independent of the issuer applying all of the tests in section 1.4 of National Instrument 43-101.

10. I have read National Instrument 43-101 and Form 43-101F1, and the Technical Report has been prepared in compliance with that instrument and form.

11.  I consent to the filing of the Technical Report with any stock exchange and other

regulatory authority and any publication by them for regulatory purposes, including

electronic publication in the public company files on their websites accessible by the

public, of the Technical Report.

Dated this 10 Day of September, 2009.

Mark D Wanless

L. Holland

Holland and Holland Consultants

9 Nevis Close,

Linslade,

Leighton Buzzard,

Bedfordshire,

England.

Telephone: (44) 1525 378294.

Email:Len@holland-holland.demon.co.uk

CERTIFICATE of AUTHOR

I, Leonard.Holland B.Sc., C. Eng., FIMM. do hereby certify that:

1.  I am a Consultant Minerals Processing Engineer of:

Holland and Holland Consultants,

9 Nevis Close,

Linslade,

Leighton Buzzard,

Bedfordshire,

England.

2.  I graduated with a degree in Extraction Metallurgy with Honours from the University of Wales in 1968.

3.  I am a Fellow of the Institute of Materials, Minerals, and Mining. UK.

4.  I have worked as a Minerals Processing Engineer for a total of 41 years since my graduation

from university.

5.  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.

6.  I am responsible for the preparation of section 16 “Mineral Processing and Metallurgical Testing” of the technical report titled “Pampa el Toro Mineral Resource Technical Report” dated September 10^th, 2009 (the “Technical Report”) relating to the mineral concentration and electric smelting of iron ore concentrate for the iron sands property. 

7.  I have had prior involvement with the property that is the subject of the Technical Report. The nature of my prior involvement was conducting pilot plant testwork on bulk samples from the mineralised resource for production of the mineral concentrate.

8.  I am not aware of any material fact or material change with respect to the subject matter

of the Technical Report that is not reflected in the Technical Report, the omission to

disclose which makes the Technical Report misleading.

9.  I am  independent of the issuer applying all of the tests in section 1.5 of National

Instrument 43-101.

10.  I have read National Instrument 43-101 and Form 43-101F1, and the Technical Report

has been prepared in compliance with that instrument and form.

11.  I consent to the filing of the Technical Report with any stock exchange and other

regulatory authority and any publication by them for regulatory purposes, including

electronic publication in the public company files on their websites accessible by the

public, of the Technical Report.

Dated this 10th Day of September, 2009.

­­­­­­­­­­­­­­­­­­­­­­­Leonard Holland. FIMMM.