Standards Were Submitted at the Rate of 1 for Every 20 Normal Samples Including Blank Material Sourced Locally

OZ Minerals Limited

Mineral Resources Explanatory Notes as at 30 June 2008

TABLE OF CONTENTS

1 MARTABE MINERAL RESOURCES STATEMENT - 30 JUNE 2008 3 2 SEPON COPPER MINERAL RESOURCE STATEMENT - 30 JUNE 2008 8 3 SEPON GOLD MINERAL RESOURCE STATEMENT - 30 JUNE 2008 27 4 PROMINENT HILL MINERAL RESOURCE STATEMENT - 30 JUNE 2008 45 5 CENTURY MINERAL RESOURCES STATEMENT - 30 JUNE 2008 56 6 DUGALD RIVER MINERAL RESOURCE STATEMENT - 30 JUNE 2008 70 7 GOLDEN GROVE MINERAL RESOURCES STATEMENT - 30 JUNE 2008 77 8 ROSEBERY MINERAL RESOURCE STATEMENT - JULY 2008 98 9 AVEBURY MINE AREA RESOURCES - JANUARY 2008 107 10 HIGH LAKE MINERAL RESOURCE STATEMENT - 30 JUNE 2008 113 11 IZOK LAKE MINERAL RESOURCE STATEMENT - 30 JUNE 2008 123

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 2 of 132 Mineral Resources 2008 Explanatory Notes

1 MARTABE MINERAL RESOURCES STATEMENT - 30 JUNE 2008

1.1 Introduction

The Martabe June 2008 Mineral Resource Statement incorporates the gold and silver Mineral Resources at the Martabe Project in North Sumatra Indonesia.

Since the previous Mineral Resource Statement in December 2007, remodelling of the Pit 1 (previously reported as Purnama) Mineral Resource has been completed. An infill program in the central part of Pit 1 between 167100N and 167300N was completed in late December 2007. A total of 35 additional holes for 6086m were drilled to infill drill the selected area using 25-meter spaced holes.

There have been no changes to the Barani (previously named Pelangi) and Ramba Joring (previously named Baskara) Mineral Resources.

1.2 Results

Table 1 Martabe Mineral Resources as at 30 June 2008

Deposit Category Tonnes Grade Contained Metal (kt) Au Ag Au Ag (g/t) (g/t) (koz) (koz) Pit 1 Measured 3,792 2.9 46 354 5,568 Indicated 47,710 1.7 22 2,604 33,513 Inferred 39,650 1.1 13 1,388 17,201 Total 91,152 1.5 19 4,345 56,282 Ramba Inferred 36,557 1.0 4 1,191 5,207 Joring Total 36,557 1.0 4 1,191 5,207 Barani Inferred 10,390 1.1 368 Total 10,390 1.1 368 Martabe Measured 3,792 2.9 354 Summary Indicated 47,710 1.7 2,604 Au ONLY Inferred 86,597 1.1 2,947 Total Mineral Resource 138,099 1.3 5,905 Pit 1 and Measured 3,792 2.9 46 354 5,568 Ramba Indicated 47,710 1.7 22 2,604 33,513 Joring ONLY Inferred 76,207 1.1 9 2,579 22,408 Au and Ag Total Mineral Resource 127,709 1.3 15 5,537 61,489

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 3 of 132 Mineral Resources 2008 Explanatory Notes

Table 2 compares the Pit 1 (Purnama) Mineral Resource at June 2008 Mineral Resources with an estimate of the tonnage and Au grade at December 2007. The Barani and Ramba Joring Mineral Resources remain unchanged.

Table 2 Comparison of Pit 1 Mineral Resource at June 2008 with December 2007 Mineral Resource Estimate

June 30 2008 Dec 31 2007 Category Tonnes Grade Au (koz) Tonnes Grade Au Au (koz) Change (kt) Au (g/t) (kt) (g/t) Au (koz) Measured 3,792 2.9 354 0 0 0 354 Indicated 47,710 1.7 2,604 48,750 1.8 2,859 -255 Inferred 39,650 1.1 1,388 42,527 1.1 1,512 -124 Total 91,152 1.5 4,34591,277 1.5 4,371 -25

The primary difference between the December 2007 and the June 2008 Pit 1 Mineral Resource is the category classification. In the new model, 4% of the tonnes have now been classified as Measured mainly at the expense of previously Indicated material, while there is also some improvement of previously Inferred material to Indicated category. Tonnages have varied slightly from the previous model largely due to changes in bulk densities, interpretations for mineralisation and oxidation zones, and parent cell sizes used in the block model.

1.3 Competent Person Statement

This Mineral Resource Statement has been compiled in accordance with the guidelines defined in the Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves (The JORC Code, 2004 Edition).

The information in this report that relates to Mineral Resources is based on information compiled by the below listed competent persons, who are Members or Fellows of the Australasian Institute of Mining and Metallurgy or the Australian Institute of Geoscientists and has sufficient experience which is relevant to the style of mineralisation and type of deposit under consideration and to the activity which they are undertaking to qualify as Competent Persons as defined in the 2004 Edition of the ‘Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves’. They consent to the inclusion in the report of the matters based on their information in the form and context in which it appears.

Deposit Competent Person Barani (previously called Iain Macfarlane, Senior Consultant Resources, Coffey Mining Ltd. Pelangi) Graham Petersen, Geology Manager Martabe, PT Agincourt Minerals Pit 1 (Purnama) and Ingvar Kirchner, Principal Resource Geologist, Coffey Mining Ltd.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 4 of 132 Mineral Resources 2008 Explanatory Notes

Deposit Competent Person Ramba Joring (Baskara) Graham Petersen, Geology Manager Martabe, PT Agincourt Minerals

1.4 Key Points relating to the June 2008 Mineral Resource Estimate

1.4.1 Drilling

The Purnama drilling database comprises an additional 6,086m of surface diamond drilling completed as infill drill holes between 167100N and 167300N, since the previous Mineral Resource statement in December 2007. The additional diamond holes are HQ and PQ diamond core drilled with LF70 rigs.

1.4.2 Drilling Quality Assurance

All of the diamond drill hole collar locations are surveyed using Total Station once completed. Down hole surveys are completed by Eastman camera with initial measurements at 20 meters from the collar with subsequent measurements at nominal 50m intervals and at the end of the hole.

Drill hole locations and orientations were validated by plotting collar pick ups against the drill design. Core recovery is generally excellent for all core sizes, averaging 92% recovery.

1.4.3 Geological Logging

All drill core is geologically logged for lithology, oxidation, alteration, structure, core recovery and RQD onto hard copy log sheets using a comprehensive set of logging codes. Zones of silic alteration and quartz veining determined during geological logging are selected for assaying.

1.4.4 Sampling

Drill core is halved and sampled at varying intervals between 1-3m and are commonly 1m sample intervals. Most drill holes are completely sampled. All samples are of diamond core.

1.4.5 Analytical Methods

All samples have undergone sample preparation at the ITS sample preparation facility in Padang prior to 250gm samples being forwarded to ITS laboratory in Jakarta for assay determination.

Since 10/7/2004 the samples are crushed to 2mm before being split to an approximate 1.5kg sample, which is pulverised to -200mesh and split to a 250 gram sample then sent to ITS Jakarta. At ITS Jakarta Au is analysed by 50g fire assay, AAS determination, Zn, Cu, Pb, Fe, Mn and Ag are analysed using HCl/HClO4 digest, Mo, As, Sb and Te by XRF, Hg by cold vapour AAS and Bi by Hydride generation with AAS finish. Most samples are analysed for Au, Ag, Cu, Pb, Zn, As and Hg only but sample suites assayed have varied over time.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 5 of 132 Mineral Resources 2008 Explanatory Notes

1.4.6 Assay Quality Control and Quality Assurance

Approximately 3-5% of pulps have been duplicated and sent to ITS Jakarta and to SGS in Balikpapan both located in Indonesia. Standards were submitted at the rate of 1 for every 20 normal samples including blank material sourced locally. Some issues of low bias to all results for the 7 Au standard reference material sourced from Geostats Pty Ltd was indicated. ITS Jakarta has indicated that this may be due to the standard matrix composition. While the consistent low bias, in the order of 1-2%, is a concern, the low bias will not significantly affect the overall result as it only applies to the recent infill drilling. As a result of this a number of matrix matched standards are being prepared using Martabe ore material and will be certified for Au and Ag. Currently Au is the only element certified for the standard material being used, although base metal standards have been used for previous Newmont data.

Blanks, duplicate pulps, blind pulp resubmissions and internal laboratory standards are also submitted and with no significant issues identified.

1.4.7 Bulk Density

Bulk density is determined on all holes at regular interval using the Archimedean method utilising plastic wrap (gladwrap) sealing of the samples. Average bulk densities were assigned in the block models according to the lithologies, oxidation and mineralised zones.

All bulk density data collected from oven dried drill core samples (0.2m lengths).

1.4.8 Interpretation

Mineral Resource modelling is carried out using DatamineTM software. The distribution of the lithologies and alteration types is well understood, but the actual control on the high grade Au mineralisation is not well understood. Interpretation of the high-grade gold mineralisation between sections in the areas of the infill drilling required some minor modification for lithology and alteration indicating reasonably good local correlation.

1.4.9 Grade Estimation

Ordinary Kriging was used to interpolate Au, Ag, Cu, AuCN, AgCN, CuCN, As and Hg grades into the block model.

Sample data were composited to 2m downhole intervals using the “residual retention” downhole compositing process, minimising the loss of information at the margins of domains and also honouring all domain boundaries. Detailed statistical and geostatistical investigations utilised the composite data. Gold and silver grade top-cuts were applied to Zone 30 of 40g/t Au and 400g/t Ag, and to Zone 40 of 30g/t Au and 250g/t Ag respectively.

1.4.10 Cut-off Grade

A lower gold cut-off grade of 0.50g/t Au was used for reporting.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 6 of 132 Mineral Resources 2008 Explanatory Notes

Extensive metallurgical testing has been completed in the past and another round of testing is underway to understand the leaching characteristics of the different styles of mineralisation and alteration with varying oxidation levels and also to determine the different mineralogical characteristics of the different ore types.

1.4.11 Mineral Resource Classification

Mineral Resources were classified based on data density, confidence in interpretation, number of samples involved in the estimate, confidence in the input data, style of mineralisation, structural and grade continuity of the mineralised zones and an understanding of some of the metallurgical aspects which may impact on successful recovery of the Mineral Resource. Based on these criteria, material has been classified as Measured, Indicated and Inferred.

Reporting is not constrained in depth or lateral extent by a nominal optimisation shell or other financial constraint.

Additional drilling and mining and metallurgical studies are required to gain further confidence in the continuity of the mineralisation, particularly at higher cut off grades.

1.4.12 Block Model Validation

Mineral Resource estimates were validated against the data used to create them both visually and statistically.

1.4.13 Recent Audits

An independent review of this Mineral Resource estimate has been conducted by Behre Dolbear Australia in June 2008.

Previous independent Mineral Resource reviews have been completed for due diligence by RSG Global now part of Coffey Mining in 2006 for Agincourt Resources and by CS-2 Pty Ltd in 2007 for Oxiana (now OZ MINERALS).

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 7 of 132 Mineral Resources 2008 Explanatory Notes

2 SEPON COPPER MINERAL RESOURCE STATEMENT - 30 JUNE 2008

2.1 Summary

The combined Measured - Indicated - Inferred copper Mineral Resource at Sepon is shown in Table 3.

Table 3 Sepon Copper Mineral Resources as at 30 June 2008.

Tonnes Grade Contained Metal (Mt) Cu Au Ag Cu Au Ag (%) (g/t) (g/t) (kt) (koz) (koz) Supergene Measured 20.9 3.38 706 Indicated 8.2 3.35 274 Inferred 31.1 1.66 517 Total 60.2 2.49 1,497 Primary Measured 2.1 1.74 0.2 7 36 13 445 Indicated 1.2 1.70 0.2 7 20 9 259 Inferred 20.2 0.97 0.3 6 197181 3,940 Total 23.5 1.08 0.3 6 253203 4,644 TOTAL 83.7 2.09 0.2 18 1,750648 27,758

Mineral Resources reported above 0.5% Cu cut-off; no further economic constraints have been for this Mineral Resource estimate

2.2 Detail

This report summarises the copper Mineral Resources as at 30th June 2008 at the Sepon Copper Mine in Lao PDR, operated by OZ LXML Sepon which is 90% owned by OZ Minerals of Australia. The copper mineralisation at Sepon is hosted in a number of deposits at varying stages of development. The supergene copper mineralisation has been estimated to contain 29.1Mt at 3.4%Cu Measured & Indicated and 31.1Mt at 1.7%Cu Inferred at a 0.5%Cu cut-off. In addition to this, 3.2Mt at 1.73%Cu Measured & Indicated and 20.2Mt at 1.0%Cu Inferred primary copper mineralisation has also been delineated at a 0.5%Cu cut-off. A summary of the changes between supergene and primary are shown in Table 4.

There has been an increase in both supergene and primary mineralisation above what has been depleted from mining at Khanong (Figure 1). During the reporting period a total of 2.8Mt (136.8ktCu) reduction occurred in the Mineral Resource during the reporting period, this was predominantly through depletion from mining at the active Khanong (KHN) copper open pit. New additions far exceeded this totalling 20.9Mt (368.4ktCu) to both the primary and supergene copper Mineral Resources during the reporting period. A new Inferred supergene and primary Mineral

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 8 of 132 Mineral Resources 2008 Explanatory Notes

Resource has been defined at Thengkham East, and additions have been defined at Thengkham South, as well as an extension of the low-grade supergene zone at Phabing. There was considerable growth in the copper stockpile from Khanong during the year, which also offset some of the reduction in the Khanong Mineral Resource. Currently gold and silver are not recovered from the processing of supergene copper mineralisation, however ongoing metallurgical test work shows that precious metals are potentially recoverable, which provides an opportunity for the extraction of precious metals from ore treated in the future.

Table 4 Sepon Copper Mineral Resource comparison 2007 – 2008.

2008 2007 Variance Tonnes Cu Tonnes Cu Tonnes Cu (Mt) (kt) (Mt) (kt) (Mt) (kt) Supergene 60.2 1497 49.3 1329 22% 13% Primary 23.5 253 16.1 189 45% 34% Total 83.7 1,750 65.5 1,519 28% 15%

Sepon Copper Resource History - Contained Metal 2,000

CuPr

1,500 CuPr

1,000

CuOx CuOx

500

0 Jun-07 Depletion Addition Jun-08

Figure 1 Sepon Cu Mineral Resource metal additions and depletions from June 2007 to June 2008 (> 0.5%Cu)

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 9 of 132 Mineral Resources 2008 Explanatory Notes

2.3 Setting

The Sepon project comprises a 1,250 square kilometres contract area located approximately 40km north of the town of Sepon, in Savannakhet Province, south-central Lao PDR. Access to the site is via sealed roads (Figure 2).

Figure 2 Sepon Location Plan

OZ LXML acquired the Sepon project from Rio Tinto in 2000. The Sepon copper operation commenced production in 2005 as a whole of ore, atmospheric leach, Solvent Extraction and Electro-winning copper processing plant. Copper mining is currently focussed on Khanong.

Figure 3 Sepon Mineral Deposit Plan

The table below follows the format of Table 1 of the 2004 JORC Code and provides a summary of important criteria related to the assessment and reporting of the copper mineralisation at Sepon.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 10 of 132 Mineral Resources 2008 Explanatory Notes

The status of each deposit between 1 July 2007 and 30 June 2008 is as follows:

• Mining, Grade Control and definition drilling during the reporting period: - Khanong (KHN)

• Mineral Resource definition drilling only during the reporting period: - Thengkham South (TKM)

- Thengkham East (TKE)

- Phabing (PHB)

• No mining or additional Mineral Resource definition drilling undertaken during the reporting period: - Thengkham North (TKN)

Supergene and primary copper has been defined for all deposits except Phabing, which features transported oxide copper mineralisation only.

Detailed internal Mineral Resource reports have been compiled for the copper Mineral Resources stated.

Criteria Status Sampling Techniques and Data

Drilling • Reverse Circulation (RC) and HQ Triple Tube Diamond Drilling (DD) has techniques been undertaken from surface to define the copper deposits. The ratio for the drilling has changed with time, for Khanong the ratio is 2:3 DD:RC, whereas for more recently drilled deposits such as Thengkham South and Phabing where current practice is to use RC methods only for drill hole precollars, the ratio approaches 2:1 DD:RC, which has resulted in improved sample quality being gained. • Down-hole surveys have been carried out using Eastman single-shot cameras or Reflex EZ tools. Surveys are taken at depths of 12m, 30m, 60m (then every 30m to the bottom of hole). Historically, vertical holes due to their short depth (~50m) have not been surveyed; these holes have generally targeted shallow dipping oxide mineralisation.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 11 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Logging • RC and DD core is logged on paper and entered manually into the database for definition drilling, data is captured digitally using pocket acQuire© for GC drilling. • Logging uses a set of pre-determined Sepon tables for; lithology, structure, mineralisation, geotech, oxidation, alteration and a site developed metallurgy code (metcode). • Logging was carried out rigorously according to the guidelines and codes presented in the March 2006 edition of the Sepon Project “Field Geologists Manual”. Drilling prior to this date has also been carried out with reference to a similar set of older logging codes that have subsequently been up-dated and compiled into the new logging reference manual. • Core is photographed and stored digitally. • All drill core is stored at the Sepon core shed.

Sub- • RC samples are collected from the cyclone and put through the riffle sampling splitter if dry. Current practice for moist and west samples is sampling by techniques quartering. A 3-5kg sample is collected for analysis. and sample • DD core is orientated along the apical trace of the reference plane (usually preparation offset 1cm from structural orientation mark when available), and then half-core samples are taken using a diamond core saw for competent core or sampling by hand using a spatula or blade for clay-rich material. • Sample interval density is measured every 10 metres using weight in air / weight in water wax immersion method. • Upon receipt at the laboratory samples are sorted, barcode tagged for tracking and weighed. The samples are then oven dried at 110°C (for core samples, minimum of 12 hours drying and for RC samples 24 hours or longer until the sample is completely dry to pass through crusher without pelleting). The entire sample is put through the Jaw Crusher with 70% passing 2mm. The sample is then rotary split and pulverised using an LM5 to 85% passing 85µm. A 110g pulp aliquot for Au Fire Assay and 20g pulp aliquot for ICP multi-element is taken. Sample preparation typically takes place at ALS laboratory Vientiane for definition drilling, and at the on-site laboratory for grade control. • The analytical procedure is as follows: - If Au grade > 10g/t, re-assayed by Fire Assay Gravimetric. - If Au grade > 0.4g/t Au, re-analysed using CN Leachwell technique. - Detection limits for Fire Assays is 0.01ppm - Ag, As, Bi, Ca, Cd, Co, Cu, Fe, Mg, Mn, Mo, Ni, P, Pb, S, Sb, Sr and Zn are analysed by ICP-OES • If Cu > 0.5%, re-assayed using an Ore Grade technique (ALS Code of (OG- 46)).

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 12 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

• The 2005-2008 Khanong Grade Control (GC) and selected infill samples were assayed at the Sepon Mine Laboratory for Cu by Tri acid digest (perchloric/hydrochloric/nitric) digest AAS determination and Au by 30g Fire Assay, AAS determination.

Quality of • All Deposits: A system whereby grade and matrix matched certified assay data standards, coarse and pulp blanks as well as field duplicates and pulp and repeats are included in the assay process for each hole. At a minimum, laboratory every hole must contain at least one coarse blank, one pulp blank and one tests standard. The guideline is that 1 in 15 samples is a QAQC control (earlier deposits have 1 in 25 or clustered 3 in 25). Rigorous checks of the laboratory results and data import procedures are undertaken regularly to identify any spurious results for verification and re-assay. Any suspect data is excluded from Mineral Resource estimation. • At Khanong, where sample analysis has been undertaken at the on-site Sepon laboratory, there is a 10-15% high relative bias in the Sepon data at grades >5% Cu (+70th percentile). However when the combined ALS+Sepon Lab Cu-population is compared with the spatially equivalent ALS lab Cu-population distribution this bias is significantly reduced (diluted). This comparison was conducted on the global spatial sub-set and on a 100m moving section-slice basis which showed that the two datasets could be combined for estimation purposes. There is no clustering of the Sepon data, suggesting that there is likely to be low risk to the Cu estimate by including this lower quality data. Further to these observations it is noted that the mineralisation domaining does not appear to be affected by the Sepon Lab data quality (the majority of this data is within the high grade CuOx mineralisation) and that the area influenced by the Sepon Lab data will be re-drilled and re-assayed as per the GC protocols for the deposit. Although not ideal, the Sepon Lab data was included in the 2008 Mineral Resource estimate dataset and it is acknowledged that, in isolation, the probability of local error may be high for Measured Mineral Resources to be estimated (particularly in the GC area), but with full consideration of all observations and variables (stated above) it is considered that the risk to the project is low and therefore Mineral Resources in the area affected by the Sepon Lab data were classified without any consideration to data quality Verification • For Mineral Resources containing large proportions of RC drilling twinned of sampling holes are routinely carried out as part of drill quality analysis and are and assaying discussed in the ‘Drill Sample Recovery’ section. • Independent / round robin laboratory checks are conducted routinely throughout the year on a quarterly or half yearly basis. The results are generally unbiased with respect to each other. In Khanong for example, the overall relative precision is ± ~6%.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 13 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Location of • Early in 2008 a LIDAR (Light Detection and Ranging) survey was data points completed providing a highly accurate topographic surface with which to facilitate the checking of locations and elevations of drill-hole collars where GPS pick-up has not been possible in heavily vegetated areas. • All drill-hole collars are converted from UTM/Indian60 projection to SPG06 local grid coordinate systems. • Drill collar locations have been validated through a process of database and spatial checking for both historical and recent data. A number of holes were identified as having suspect locations. These issues were resolved prior to modelling of the data. Data spacing • Drill hole spacing generally ranges from 100x50m to 50x50m in Inferred and zones, 25x25m in Indicated zones and where GC has been undertaken at distribution 7.5x7.5m in the Measured zones. Orientation • The bulk of copper oxide mineralisation exists in sub-horizontal domains of data in within a supergene zone. The lower Cu-carbonate/oxide domain exists as relation to thin blankets. These zones are best targeted with near vertical drill holes, geological especially at Khanong. structure • Drill holes into some the copper mineralised domains at Thengkham are a mixture of vertical to 60º dipping holes orientated either N-S or E-W depending on the structure and/or stratigraphic control being targeted, especially with reference to primary copper mineralisation, which has some structural control. • Analysis suggests there is strong and consistent orientation to the mineralisation within some domains (esp. TKN & TKS) that is either better sampled or over-sampled (bias) by the E-W drilling compared to the N-S (and sub-vertical) drilling, which goes against the overall strike of the deposit. Audits and • AMC consultants conducted a review of the 2008 Phabing and reviews Thengkham South Mineral Resource models which were updated during the reporting period. • All other models have been subject to a series of internal and external reviews during their history of development.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 14 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Estimating and Reporting of Mineral Resources Database • The Sepon geological database system consists of three components. A integrity manual Field Logging System, a Data Entry Database (DEDB), and a Master Database (LaoDB). Each digital component is configured to run in SQL Server with user access and permissions. • The DEDB works as a quarantine and compilation system. The supervising database geologist reviews all new data against original paper logs, with corrections made prior to loading into LaoDB which is done via SQL Server stored procedures to detect and hold any errors on import. © • The GBIS database and logging system was introduced in 2006 and populated from the existing acQuire database. Ongoing analytical data is uploaded directly from laboratory SIF files. • A recent external audit (by IO Global Ltd Pty) of the database found that the post-2006 analytical data to be of appropriate integrity. Geological • KHN - The bulk of copper and silver mineralisation at Khanong exists in interpretation sub-horizontal domains within a supergene blanket in the upper reaches of Khanong creek. The lower Cu-carbonate/oxide domain exists as thin blankets (5m – 10m thick) overlaying fresh dolomitic footwall lithologies. It is predominantly high grade and contains approximately 32% of the 2008 copper Mineral Resources (1% Cu cut). 47% of the 2008 copper Mineral Resources are contained in two overlaying chalcocite clay horizons that are up to 50m thick in places. The remainder of the supergene zone consists of low-grade limonitic and “clean” clays and a surface gossan. Copper mineralisation also exists in three shallow dipping fault zones in the north of the deposit and in primary sulphides in the south of the deposit. All of these domains were modelled for Mineral Resource estimation. A satellite chalcocite mineralized domain identified in 2005 exists to the south of the main deposit and is located within a narrow south-west striking depression in the supergene to fresh-rock contact.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 15 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status • TKE/TKN/TKS - The Thengkham district contains a zoned system of primary copper, gold, silver and molybdenum mineralisation which is believed to have formed late in the intrusive history. The primary mineralisation ranges from central porphyry style Mo-Cu-Ag mineralisation through to retrograde skarn Cu-Mo-Au-Ag-Bi mineralisation associated with carbonate rocks close to the intrusive centres. Weathering processes have resulted in the formation of secondary supergene copper and gold deposits. The bulk of supergene Cu and associated Au-Mo mineralisation exists as moderately dipping tabular to flat zones within the weathering profile. The continuity of mineralisation has varying strike lengths along the northern and southern slopes of the E-W striking Thengkham ridge. Chalcocite supergene mineralisation is located adjacent to primary pyrite / chalcopyrite mineralisation. Copper oxide and carbonate mineralisation is best developed further down slope and the Cu in this mineralisation has been remobilised during the weathering process. • PHB - Phabing is interpreted to be a transported supergene copper mineralisation from the slopes of Thengkham North, and has no primary copper directly associated with the deposit. There are major E-W bounded fault zones that are interpreted to control two distinct styles of mineralisation on either side of the fault. Mineralisation on the north side of the fault zone is controlled by a large dolomite unit. Copper mineralisation is highest around the contact with fresh dolomite. The dolomite rock mass is part of the Nalou Fm; which is folded as part of a syncline structure. Mineralisation on the south flank of the fault is most developed around the contact with calc-shale units. These areas have all been domained separately

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 16 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Dimensions Project Min Max Dimension Number X 26,750 29,150 25 96 KHN Y 75,000 76,500 25 60 Z 150 650 5 100 X 16,350 20,000 25 146 TKN Y 73,200 76,000 25 112 Z 0 600 5 120 X 16,430 20,850 20 221 TKM Y 73,160 74,720 12 130 Z 40 600 2.5 224 X 20,775 21,275 50 10 TKE Y 74,460 75,260 50 16 Z 50 450 25 16 X 15,950 17,730 10 170 PHB Y 74,250 75,462 6 202 Z 0 500 2.5 200 Estimation KHN - A summary of the estimation used at Khanong is: - Block model panel sizes reflected the drill hole and sample spacing, and domain morphology with parent blocks in the definition drilling grid being 25x25x5m (E, N, RL) and sub-blocking allowed to 5x5x1m in size. Parent blocks within the GC drilling grid are 12.5x12.5x2.5m in size with sub-blocking allowed to 2.5x2.5x0.5m in size. - 2m drill-hole composites were selected following down-hole variography studies and grades were estimated into domains using OK and the VulcanTM mining software. - The deposit was divided into two search ellipsoid-orientation zones, reflecting the change in orientation of the supergene blanket, and variography and grade estimation was conducted for each of the mineralized domains in these zones. - Three search ellipsoids and sample moisture criteria were used for composite selection and classifying the Mineral Resource estimate. -

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 17 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status - Thorough statistical analysis of the geology and assay data identified that the copper mineralisation at Khanong required, and could be domained for grade estimation. Sharp contacts and differences in characteristic features such as copper grade ranges, observed mineral species, colour, clay content, alteration and secondary element abundance can be observed between the modelled domains. Mineralisation continuity across the deposit enabled the application of surface modelling techniques (MinesightTM software) which were later combined to form solids used in the block modelling process. The 2007 upper and lower surfaces of modelled domains were edited to incorporate the additional 2008 data and create the solid models used in the 2008 Mineral Resource Estimate. • TKE - A summary of the estimation used at Thengkham East is: - Block model panel sizes reflected the drill hole and sample spacing, and domain morphology with parent blocks in the definition drilling grid being 50x50x25m in size (East, North, RL) and sub- blocking allowed to 5x5x5m in size. - 1m drill-hole composites were selected on the basis of summary statistics and were used for variogram analysis and grade estimation. - Variogram analysis was undertaken using IsatisTM statistical software. Grades were estimated into hard boundary domains using ordinary kriging and VulcanTM mining software. Variables interpolated include Cu, Au, Ag, Mo, S, Mn and Fe. Search ellipse orientations were determined by geology domain shape and variogram analysis. - Each domain was estimated using two passes with increasing search ellipse size (pass 1 - 100:60:30m and pass 2 – 200:120:60m). Minimum number of composites per estimate was 5 with a maximum of 20. Upper cuts were selected on the basis of statistical analysis and are above the 97.5%.

• TKN - A summary of the estimation used at Thengkham North is: - Grade was interpolated into blocks with a parent cell size of 25m x 25m x 5m (X, Y, Z). The block size was selected based on drill spacing in the more sparsely drilled areas and because these dimensions are appropriate for production models currently in use on site.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 18 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status - Mineral Resource drilling (both RC and DD) is nominally sampled at 1m intervals. A 2m composite was chosen for the Thengkham North statistical analysis and grade estimation after looking at a range of composite lengths from 1m to 5m. The selection of 2m composites took into consideration grade statistics and the bulk mining techniques being used at Sepon. - The geostatistical analysis was undertaken by AMC consultants using Isatis© geostatistical software package. From this work variography and estimation parameters were developed for Cu-Au- Ag-Mo-S-Mn-Fe-Ca-Mn-Al and pyrite. - A grade upper-cut did not make a significant difference to the global statistics for each of the modelled domains except for domains with low numbers of composites. Where groups of high grades were grouped together they generally were not considered as outliers. - The relative nugget effect ranges from 10-25%. The average drill hole spacing is ~ 25m with the majority of the semi-variogram variance consumed by ranges with values between 20-40m. The longest range structures generally vary between 100-250m. Shortest range structures are generally in the vertical direction contained by the densest composite data. - Copper grades were estimated into blocks by Ordinary Kriging (OK) using Datamine software. Two search pass were implemented to interpolate a complete model. The first search pass filled all of all cells with copper values. The second search pass, which was twice the size of the first pass, was required for some minor elements. Additional quality estimators were informed by two passes; if uninformed they were left absent. - A minimum of four composites a maximum of 10 were used to inform a cell with grade for the first search pass. The second search pass used a minimum number of 3 samples and maximum number of 10 samples. - CUO - Copper carbonate domains: Cu greater than or equal to 1%Cu; Predominant rock codes of CUO, with lesser LCC; Oxidation logged as >1; Sulphur grade < ~ 0.5%S. - MNO - Manganese oxide-rich clay domains: Cu grade generally between 0.2% - 1.0%Cu; Mn grade generally above 0.2%Mn; Fe grade < ~ 15%; S grade < 0.5%; oxidation logged as >1; Metcodes used to constrain are MNO and LCC.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 19 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status - CC - Chalcocite: Oxidation logged as >1; S > ~ 0.2%, logged code = CC, Fe ~ Cu + S, if Fe >> Cu + S then Metcode = PY; logged Metcode = PY, Cu >1% and minor oxidation; Metcode ≠ CC, but high S and high Cu. - PY - Primary Copper: All Cu mineralisation above ~ 0.2%Cu below Base of Supergene Mineralisation (BOSM), pyrite / chalcopyrite mineralisation; High Fe:S ratio. • TKM - As for TKN with minor modifications as follows: - Block size 25x10x5m (E, N, Rl) - Elements modelled Cu-Au-Ag-Mo-S-Fe-Mn-As-Pb-Zn-Ca-Mg-Al. With density assigned by oxide state in waste and estimated by nearest neighbour within mineralised domains, and pyrite calculated by empirical means. - Search parameters were provided by AMC consultants, and included minimum samples 4 – maximum 14. - Top-cuts for estimation were typically the 98-99th percentile. - Block discretisation was 10:4:2 (E, N, RL) - Search Ellipse typically ranged from 150-340 (Major) – 90-150 (Semi-Major) – 50-150m (Minor) for copper. • PHB - A summary of the estimation used at Phabing is: - A block size of 10mX x 6mY x 2.5mZ was used for this estimate. This block size fits in with the standard block size which has been used for copper estimates, where these dimensions represented the best compromise between the drill spacing, spatial continuity and anisotropy of mineralisation, the sub-horizontal nature of most of the domains, estimation precision and the proposed scale of mining. - The compositing and subsequent data processing and statistical analysis were carried out in Minesight Compass™ software. The composite length used for Mineral Resource definition data was 2.0m, or twice the default sample length. This was chosen as the length which best honoured the majority of sample intervals, the dimensions of geological domains, and the bench height (2.5m) or an alternative statistical reason. Composites were created honouring the estimation domain coding, and storing the coordinates of composite midpoints. The option to merge up small intervals at the end of zones into the previous interval when the interval is shorter than half the composite interval was used. Consequently the last composite generated in a domain can range between 1.0 and 3.0m in length. Cu, Mn, Au, Ag, Mo and S values were composited.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 20 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status - Sectional interpretation of mineralised zones were constructed using digitised and representative ‘2-D’ sectional mineralised zone ‘strings’. This was first done for the copper mineralisation as interpreted from drill-hole intercepts. - Copper, manganese, gold, silver, molybdenum and sulphur values have been interpolated using OK in Minesight™ software. An OK estimate is considered to be an appropriate input for the purposes of reporting global and local Mineral Resources. Estimation was performed in a single pass, because most domains have been sampled on a regular grid pattern and samples are thus similarly located with respect to blocks. In addition to grade, the number of informing samples, the distance to the closest sample and the kriging slope of regression were stored to the block. - Three MNO and seven CUO domains were estimated. Each was estimated with a minimum of 4 samples and maximum of 16. Major-Semi-Minor ranges in the search ellipse vary within each domain and were determined by geostatistical analysis from AMC consultants and review of the mineralised domain geometry. th • Top-cuts were restricted to the 97-99 percentile where applied. Modelling • KHN - is an update of the 2007 OK block model and includes data from techniques additional, in-pit mapping, grade control, geotechnical and infill drill holes. • TKE - is an OK block model created from historical & exploration drilling during the reporting period. • TKN - is an OK block model and remains unchanged from 2007. • TKM - is an OK block model, updated from the 2007 MIK model, includes data from infill drill holes • PHB - is an OK block model created from definition drilling. Moisture • Moisture has been removed from all tonnage estimates. Cut-off • The final Mineral Resource estimate is reported at a 0.5% Cu cut-off and parameters further tabulated and sub-divided by material oxidation state and Mineral Resource classification category. Mining • No mining assumptions have been made. assumptions • However, mining at Sepon is currently undertaken using open pit methods. • Some deeper primary Mineral Resources may be more appropriately accessed by underground methods.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 21 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Bulk density • Samples for bulk density determination are taken from diamond drill core every 10m using a wax immersion method. Twenty centimetre samples are dried for 24 hours in a drying oven. The samples are then weighed as dry samples. They are then coated in wax and weighed in water and then in air. The specific gravity of the wax is also measured to define any differences in specific gravity due to the addition of the wax coating. • For most deposits bulk densities were estimated as a variable within the block model within the mineralised zones based on statistical analysis of the data, and consideration of the host lithology. Background waste is assigned a designated SG value from statistical analysis on the basis of its oxidation state. • Khanong: - 889 wax-immersion bulk density determinations were collected from diamond drill holes across the deposit (pre 2002 drilling) and form the basis for assigning bulk densities to the Mineral Resource estimate. Average densities were applied to the copper- carbonate/oxide domain and the structurally controlled mineralisation. Densities were assigned to the chalcocite clay, limonitic clay, gossan and primary sulphide domains by regression with either clay percentage and/or Fe grade. Validation by way of averaging the modelled/assigned bulk densities and the raw data for each domain showed excellent correlation.

Metallurgical • A series of Metcode domains have been developed for the western factors or copper deposits (TKE, TKM, TKN, PHB) that include: assumptions - CC – Chalcocite-clay - CUO – Copper oxide - LCC – Limonitic clay - LIM – Limonite ironstone - LSO – Limonitic soil - MNO – Manganese oxide - OX – Oxide - POX – Partial oxide - PR – Primary - PY – Pyrite-chalcopyrite

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 22 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status For Khanong: - Goss – Gossan - SRC – Limonitic Supergene Clay - CC – Chalcocite Clay - Cu-CARB – Copper Carbonate/Oxide - Below-SRC – Undifferentiated Footwall - Py-MSS – Primary Sulphide These codes are based on the metallurgical behaviour in the Sepon processing plant. • The elements estimated for most mineral deposits were Cu, Ag, Au and Mo for economic assessment; Al, Ca, Fe, Mg, S and calculated pyrite for plant design considerations; and Al, As, Ca, Fe, Mg, Mo and S for environmental studies. In addition the model was coded by geometallurgical type for scheduling the process plant and by lithology for determining mining costs and geotechnical domains. • Khanong variables estimated were Cu, Au, Ag, Fe, S and Calc_Py. • Thengkham East (TKE) variables included Cu, Au, Mn, Mo, S and Fe. Classification • KHN - Classification is based on the following: - Run 1: 70m x 70m x 6m ellipse (E, N, RL directions), octant search with four octants informed, minimum composite number of 10 and maximum of 20. Unless otherwise stated blocks informed with copper grades from this run are assigned the Measured category. - Run 2: 100m x 100m x 9m ellipse – octant search, with four octants informed, minimum sample number of 8 and maximum of 30 and a restriction on extreme grade influence is applied. Unless otherwise stated blocks informed with copper grades from this run are assigned the Indicated category. - Run 3: 120m x 120m x 16m ellipse – non-octant search, minimum sample number of 4 and maximum of 30 and a restriction on extreme grade influence is applied. Unless otherwise stated blocks informed with copper grades from this run are assigned the Inferred category - The block model and grade estimates were validated using statistical and visual methods with good reconciliation established between the estimated grades and the composited drilling data. Reconciliation between the 2002 and subsequent Mineral Resource models shows that later models differ only where additional data exists. 99kt of copper metal have been added to the 2002 Mineral Resource estimate over these years (0.5%Cu cut).

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 23 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

• TKN – Mineral Resource classification is based solely on confidence in copper grade and volume (tonnage) estimation. - Three dimensional contour surfaces (shells) were created at appropriate values for the following criteria; sample moisture, sample recovery, sample type, relative Kriging Variance, Kriging slope of regression, number of samples used to inform the cell with grade and the number of search passes required to fill a cell with a gold grade. - These shells were then assessed on a domain by domain basis in conjunction with a subjective assessment of the confidence in the interpretation of geological controls on mineralisation and the spatial configuration of drilling within the domain. - No Mineral Resources are reported from the four waste domains as the geological and grade continuity of potentially economic intersections within these domains has not been demonstrated. - Only Inferred Mineral Resources are reported from the manganese clay and fresh domains as the geological and grade continuity of potentially economic intersections within these domains has not been demonstrated to a sufficient level of confidence within these two domains. - No Measured Mineral Resources have been reported from Thengkham North as the current 25 m by 25 m drilling pattern does not allow sufficient confidence in either geological interpretation or grade estimation. • TKM - Quality indicators were applied based on the following: - Distance to nearest composite (m) & number or composites - Kriging Variance - Mineralised Domain Classification was then applied on a block-by block basis, and subsequently reviewed against the geological confidence and supporting data following the model validation process. • TKE - Entirely classified as Inferred on the basis of drill hole spacing and geological confidence. • PHB - Quality indicators were applied based on the following: - Distance to nearest composite (m) - Number or composites used - Kriging Variance - Mineralised Domain

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 24 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status • Classification was then applied on a block-by block basis, and subsequently reviewed against the geological confidence and supporting data following the model validation process.

Audits and • OZ Minerals has an internal ‘Mineral Resource and Ore Reserve Policy’ reviews that requires at a minimum external reviews every two years, and internal company review every interim year by a sub-set of the OZ Minerals Resource and Reserve Committee. LXML undergoes a rigorous internal peer review at the completion of every model update.

2.4 Mineral Resource Estimation Contributions

The following identifies the key personnel involved in the estimation of the Sepon Copper Resources:

KHN – All grade control data was sourced and validated by LXML geologists. Geological interpretations performed by Duncan Hackman of Hackman and Associates Pty Ltd (HA). Geostatistical analysis was conducted by Arnold van der Heyden of Hellman and Schofield Pty Ltd (HS) and grade interpolations were performed by HA.

TKN – Drill hole data was sourced, validated and the geological interpretation and domaining were undertaken by LXML geologists and Kerrin Allwood of Geomodelling Pty Ltd (GM). Geostatisitical analysis, Mineral Resource modelling and grade interpolation were conducted by Mark Sweeney and Tracie Burrows of AMC Consultants (AMC).

TKS – Drill hole data was sourced and validated by LXML geologists and Kerrin Allwood of Geomodelling Pty Ltd (GM). Geological interpretation, domaining of the data and grade interpolation was performed by LXML geologist Linda Sprigg. Geostatisitical analysis was conducted by Steve Hyland of Ravensgate Pty Ltd (RG) for Primary Copper, and Mark Sweeny of AMC consultants (AMC) for Oxide Copper.

TKE – Drill hole data and geological interpretation was compiled and validated by James Cannell (LXML). Domaining of data, grade interpolation and geostatistical analysis was conducted by Jared Broome (OZ Minerals).

PHB – Drill hole data and geological interpretation was complied and validated by LXML geologists. Geostatistical analysis was conducted by Mark Sweeny of AMC. Domaining of data and grade interpolation was performed by LXML geologist Koko Suhanto.

2.5 Competent Person Statement

This Mineral Resource Statement has been compiled in the accordance with the guidelines defined in the Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves (The JORC Code, 2004 Edition).

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 25 of 132 Mineral Resources 2008 Explanatory Notes

The information in this report that relates to Mineral Resources is based on information compiled by Jason McNamara and Terry Briggs under the supervision of Paul Quigley a full-time employee of OZ Minerals in the position of Manager - Geology, OZ LXML Sepon, and is a member of the AusIMM and AIG.

Paul Quigley BSc (Geol), MSc, MAusIMM, MAIG, RPGeo, full-time employee of OZ Minerals, has over 20 years experience as a geologist in mining and exploration, which includes 11 years experience in resource estimation and 8 years experience in sediment-hosted gold deposits. He also has substantial knowledge of the matters relating to supergene copper estimation, mine production, and reconciliation, and as such is responsible for data integrity, geological interpretation and resource classification.

Duncan Hackman B.App.Sc (Geol), MSc, MAIG, employed by Hackman and Associates Pty Ltd, has over 23 years experience as a geologist in mining and exploration, which includes 16 years experience in resource estimation, 7 years experience in supergene copper deposits.

Jared Broome B.App.Sc (Geol)(Hons), MSc, Grad. Dip. Mining, Grad. Dip. Business, FAusIMM, MAIG, full-time employee of OZ Minerals has over 16 years experience as a geologist in mining and exploration which includes 13 years experience in resource estimation of base and precious metal deposits.

Competent Person responsibility is as follows:

TKN, TKS & PHB – Paul Quigley (LXML). KHN – Duncan Hackman (HA). TKE – Jared Broome (OZ Minerals)

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 26 of 132 Mineral Resources 2008 Explanatory Notes

3 SEPON GOLD MINERAL RESOURCE STATEMENT - 30 JUNE 2008

3.1 Summary

Table 5 Sepon Gold Mineral Resources as at 30 June 2008.

Tonnes Grade Contained Metal (Mt) Au (g/t) Ag (g/t) Au (koz) Ag (koz) Oxide Measured 7.9 1.1 4.9 278 1259 Indicated 4.7 1.6 4.8 240 716 Inferred 5.4 1.0 4.1 168 713 Total 18.0 1.2 4.6 686 2688 Partial Oxide Measured 0.6 3.0 11.0 55 199 Indicated 4.0 2.5 8.5 319 1103 Inferred 2.2 1.1 4.7 79 334 Total 6.8 2.1 7.5 453 1635 Primary Measured 5.1 2.9 7 479 1184 Indicated 13.2 2.7 8 1132 3505 Inferred 6.1 1.8 7 352 1457 Total 24.4 2.5 8 1963 6153 TOTAL 49.2 2.0 7 3104 10488 Mineral Resources reported above 0.5g/t Au cut-off for Oxide and Partial and 1.0g/t Au for Primary.

No further economic constraints have been applied for this Mineral Resource estimate.

3.2 Detail

This report summarises the gold oxide Mineral Resources as at 30 June 2008 at the Sepon Gold Mine in Lao PDR, operated by OZ LXML Sepon which is 90% owned by OZ Minerals of Australia. The gold mineralisation at Sepon is hosted in a number of deposits at varying stages of development.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 27 of 132 Mineral Resources 2008 Explanatory Notes

The gold mineralisation has been estimated to contain 49.2Mt @ 2.0g/t Au and 7g/t Ag in total (Measured, Indicated, Inferred). During the reporting period a total of 1.4Mt (73.6koz) was depleted from the previously reported Mineral Resource (Figure 4), which is largely attributed to depletion through mining of the gold oxide pits, additional processing from previously developed stockpiles, and a negative correction to the Houay Yeng Mineral Resource through additional resource definition drilling. New additions of 1.3Mt (42.7koz) were made to the gold oxide Mineral Resource at Phabing, Thengkham South, Thengkham East and Khanong. The primary gold has been reported above a 1.0g/t Au cut off, which has been increased from previously estimates reporting above a 0.5g/t Au cut off.

Sepon Gold Resource History - Contained Metal 3500

3000

2500

2000 AuPr AuPr koz

1500

1000

500

AuOx AuOx

0 Jun-07 Depletion Addition Jun-08

Figure 4 Sepon Au oxide Mineral Resource contained metal deletions and additions from June 2007 to June 2008 (> 0.5g/tAu)

3.3 Setting

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 28 of 132 Mineral Resources 2008 Explanatory Notes

Figure 5 Sepon Location Plan

OZ LXML acquired the Sepon project from Rio Tinto in 2000. Gold has been produced at Sepon since December 2002. In early 2005, an expansion of the original gold processing facility was completed doubling the capacity of the gold processing plant, to 2.5Mt/a. Production is currently focussed on Discovery, Houay Yeng and Pha Vat North (Figure 6).

Figure 6 Sepon Mineral Deposit Plan

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 29 of 132 Mineral Resources 2008 Explanatory Notes

3.4 Assessment and Reporting Criteria

The following table provides a summary of important criteria related to the assessment and reporting of the gold oxide mineralisation at Sepon.

The status of each deposit between 1 July 2007 – 30 June 2008 is as follows:

• Mining and Grade Control drilling during the reporting period: - Khanong Au-Gossan (KHN-Au) • Mining only during the reporting period: - Discovery East Oxide (DSE) - Namkok East Oxide (NKE) - Nalou Oxide (NLU) • Mineral Resource definition drilling during the reporting period. No mining: - Thengkham East (TKE); Thengkham South (TKM) - Phabing (PHB) - Houay Yeng (YNG) • No mining or additional work undertaken during the reporting period: - Discovery Main (DIS); Discovery Colluvial (DSC); Discovery West (DSW) - Dankoy (DKY) - Muang Luang (LOL) - Namkok West (NKW) - Phavat North (PNV) - Thengkham North (TKN) - Vang Ngang East (VNE)

Detailed internal Mineral Resource reports have been compiled for 2008 update of Houay Yeng, and extensive JORC check-lists have been created for internal documentation for previous Mineral Resource updates.

The Khanong and Thengkham East mineral deposits have estimated gold as a secondary element of economic interest to copper. Gold Mineral Resources for these deposits above a 0.5g/tAu cut off and below 0.5% Cu has been reported as part of the total Sepon Gold Mineral Resource. The gold Mineral Resources at Khanong are not currently planned to be recovered and the Mineral Resource classification category has been kept the same as for the copper Mineral Resources. Further estimation details for Khanong and Thengkham East are provided in Section 2, Sepon Copper Mineral Resource Statement - 30 June 2008.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 30 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Sampling Techniques and Data Drilling • Reverse Circulation (RC) and Diamond Drilling (DD) has been undertaken. techniques For historical deposits, as much as 80-90% is defined by RC drilling (resource definition or grade control - GC). For recent deposits this figure is closer to 60%, with DD undertaken using HQ triple tube. • Holes greater than 50m were routinely down hole surveyed through the drill rods (chrome rods) for RC holes. It is judged that sample location errors on shorter holes (i.e. Less than 50-70m) are immaterial as holes are generally at high angles to mineralisation geometries. In effect, missing survey data are unlikely to significantly affect Mineral Resource estimates. Drill sample • Sample recoveries tend to be better in DD (90%) than RC (70% - recovery calculated) with minor differences between mineralisation and waste rock. Sample recovery is better in primary rock than transitional and oxide as expected. There tends to be a bias towards RC, especially early drilling when wet samples were encountered, the practice in latter drilled deposits is to favour towards DD in this situation. Recoveries tend to be marginally lower in mineralised zones, but from the analysis of twinned holes, shows that this is expected to have a negligible impact on the estimation. Logging • RC and DD core is logged on paper and entered manually into the database for definition drilling, data is captured digitally using pocket acQuire© for Grade Control (GC) drilling. • Logging uses a set of pre-determined Sepon table for; Lithology, Structure, Mineralisation, Geotech, Oxidation, Alteration and a site developed Metcode. • Core is photographed and stored digitally. • All drill core is stored at the Sepon core shed. Sub- • RC samples are collected from the cyclone and put through the riffle sampling splitter if dry. Current practice for moist and wet samples is sampling by techniques quartering. A 3-5kg sample is collected for analysis. and sample • DD core is orientated along the apical trace of the reference plane (usually preparation offset 1cm from structural orientation mark when available), and then half-core samples are taken using a diamond core saw for competent core or sampling by hand using a spatula or blade for clay-rich material. • Sample interval density is measured every 10 metres using weight in air / weight in water wax immersion method.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 31 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

• Upon receipt of sample at the laboratory samples are sorted, barcode tagged for tracking and weighed. The samples are then oven dried at 110°C (for core samples, minimum of 12 hours drying and for RC samples 24 hours or longer until the sample is completely dry to pass through crusher without pelleting). The entire sample is put through the Jaw Crusher with 70% passing 2mm. The sample is then rotary split to 3kg if required then pulverised using an LM5 to 85% passing 85µm. A 110g pulp aliquot for Au Fire Assay and 20g pulp aliquot for ICP multielement is taken. Sample process typically takes place at ALS laboratory Vientiane for definition drilling, and at the on-site laboratory for grade control. • The analytical procedure is as follows: - If Au grade > 10g/t, re-assayed by Fire Assay Gravimetric. - If Au grade > 0.4g/t Au, re-analysed using CN Leachwell technique. - Detection limit for Fire Assay is 0.01ppm - Ag, As, Bi, Ca, Cd, Co, Cu, Fe, Mg, Mn, Mo, Ni, P, Pb, S, Sb, Sr and Zn are analysed by ICP-OES - If Cu > 0.5%, re-assayed using an Ore Grade technique. Quality of • A system whereby grade and matrix matched certified standards, coarse assay data and pulp blanks as well as field duplicates and pulp repeats are included and in the assay process for each hole. At a minimum, every hole must laboratory contain at least one coarse blank, one pulp blank and one standard. The tests guideline is that a minimum of 1 in 15 samples is a QAQC control (earlier deposits range from 1 in 25 to clusters of 3 in 25). Rigorous checks of the laboratory results and data import procedures are undertaken regularly to identify any spurious results for verification and reassay. Any suspect data is excluded from Mineral Resource estimation. Verification • For Mineral Resources containing large proportions of RC drilling twinned of sampling holes are periodically drilled as part of drill quality analysis and are and assaying discussed in the ‘Drill Sample Recovery’ section. The general conclusion is that wet and moist samples (RC) have demonstrated smearing and a positive grade bias. As a consequence, these samples are given a lower level of confidence in the data when used for evaluation, and current practice is to use DD when wet conditions are experienced. • Independent / round robin laboratory checks are conducted on a quarterly or half yearly basis. The results are generally unbiased with respect to each other, for example, Khanong has an overall relative precision of +/- ~6%.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 32 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Location of • Early in 2008 a LIDAR (Light Detection and Ranging) survey was data points completed providing a highly accurate topographic surface with which to facilitate the checking of locations and elevations of drill hole collars where GPS pick-up has not been possible in the heavily vegetated terrain. • All drill-hole collars are converted from UTM/Indian60 projection to SPG06 local grid coordinate systems. • Drill collar locations have been validated through a process of database and spatial checking for both historical and recent data. A number of holes were identified as having suspect locations. These issues were resolved prior to modelling of the data. Data spacing • Drill hole spacing generally ranges from 100x50m to 50x50m in Inferred and zones, 25x25m in Indicated zones and where grade control has been distribution undertaken at 10x5m in the Measured zones. Orientation • Most drill holes are drilled with dips of -60º from horizontal to intersect of data in steeper structures, or vertically for flatter dipping stratigraphically relation to controlled mineralisation. Where possible, drilling is oriented geological perpendicular to the mineralised zones. structure • Drill hole orientation and depths were checked against site generated cross-sections. Audits and • AMC conducted a review of the 2008 Houay Yeng Mineral Resource reviews model which was updated during the reporting period. • Historical models have all been subject to a series of internal and external reviews during their history of development. Estimating and Reporting of Mineral Resources Database • The Sepon geological database system consists of three components. A integrity manual Field Logging System, a Data Entry Database (DEDB), and a Master Database (LaoDB). Each digital component is configured to run in SQL Server with user access and permissions. • The DEDB works as a quarantine and compilation system. The supervising database geologist reviews all new data against original paper logs, with corrections made prior to loading into LaoDB which is done via SQL Server stored procedures to detect and hold any errors on import. © • The GBIS database and logging system was introduced in 2006 and populated from the pre-existing aQuire© database. Ongoing analytical data is uploaded directly from laboratory SIF files. • A recent external audit (IO Global) of the database found that the post- 2006 analytical data to be of appropriate integrity.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 33 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Geological • The majority of Sepon gold deposits are accompaniments to interpretation decalcification and silicification of calc-siltstone, either along faults or preferred lithological contacts. Contacts of calc-siltstone with dolomite, non-calcareous siltstone, chert and rhyodacite porphyry are preferentially mineralised. Massive jasperoid is widely mineralized and typically contains the highest-grade gold mineralisation. • Supergene oxidation of the primary refractory sediment-hosted gold mineralisation is critical to the characterisation of mineralisation due to its ability to destroy both gold-bearing pyrite and activated organic carbon. Hence, rock permeability is an important factor in the characterisation of the gold mineralisation and its control on oxidation depth. • Larger scale distribution of gold mineralisation is controlled by the intersection of structures and reactive stratigraphic intervals. The geometry of the controlling structures and known reactive host rocks is well constrained in the area of known Mineral Resources. In primary mineralisation, most of the gold occurs as micro-disseminations associated with overgrowths of pyrite on early pyrite grains. • A significant component of the oxide mineralisation is a supergene karst- controlled residual facies of mineralisation. In this case the position of the primary mineralisation sits above a carbonate unit in the Nalou Formation. Oxidation of the sulphides in the mineralised rock causes dissolution and karstification of the underlying carbonate and the insoluble material, including gold is trapped and concentrated in the karst. Supergene enrichment of gold grade occurs in this process. • Unlike the other gold deposits, Houay Yeng is interpreted to be part of the Vang Ngang Formation. The main lithological units include an interbedded siltstone and sandstone, limestone and a siliciclastic sequence of siltstones and cherts, with locally intruded rhyodacite porphyry sills. There is a significant steeply dipping structure through the deposit. Most of the mineralisation is associated within the karst infill collapse breccia that has formed at the boundary of the limestones and overlying siltstones / cherts. • The gold mineralisation is slightly different for the copper deposits (reported >0.5g/tAu <0.5%Cu). At Thengkham, a zone of enhanced gold and manganese grades occurs above and in the upper parts of the supergene copper mineralisation, at Thengkham South and East this is a limonite zone. The Au-Mn zone is in turn overlain by a near surface iron oxide rich zone which in places has sufficient iron to be called a gossan. At Khanong, the gold is more elevated within the gossan zone.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 34 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Dimensions Project Min Max Dimension Number X 24,500 28,175 25 108 Sepon site NKW NKE Y 74,000 75,278 6 213 grid SPG06 Z 122.5 270 2.5 59 X 27,850 28,850 25 40 VNE Y 73,120 73,510 6 65 Z 100 330 2.5 92 X 22,700 24,500 15 120 NLU Y 73,400 75,278 6 313 Z 62.5 300 2.5 95 X 25,500 26,550 15 70 LOL Y 76,600 77,200 6 100 Z -400 350 2.5 300 X 22,600 28,000 5 216 DIS DSW Y 73,882 77,002 3 130 NLU Z -30 500 1.25 212 X 27,700 28,175 25 25 DIS-F Y 75,250 76,900 6 6 Z -150 450 2.5 2.5 X 25,500 26,550 15 70 PVN Y 76,600 77,200 6 100 Z -400 350 2.5 300 X 26,750 29,150 25 96 KHN – Au Y 75,000 76,500 25 60 Z 150 650 5 100 X 16,430 20,850 20 221 TKM - Au Y 73,160 74,720 12 130 Z 40 600 2.5 224 X 20,775 21,275 50 10 TKE – Au Y 74,460 75,260 50 16 Z 50 450 25 16 TKN - Au X 16,350 20,000 25 146 Y 73,200 76,000 25 112

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 35 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Z 0 600 5 120 X 15,950 17,730 10 170 PHB - Au Y 74,250 75,462 6 202 Z 0 500 2.5 200 X 18,700 19,975 15 85 YNG Y 71,700 72,420 6 120 Z 50 400 2.5 140 Estimation Oxide: DIS, DSC, DSW, NLU, NKE, NKW, VNE, PVN, LOL, YNG

• A composite length of 2m down hole was chosen, and the compositing process checked and validated. • Variography and search neighbourhood optimisation for each domain was performed using a combination of Minesight© and Isatis© geostatistical software packages. In general, the spatial continuity of gold grades shows a moderate nugget effect ratio, and short-moderate ranges of spatial continuity. • Gold grades were estimated into blocks by ordinary kriging (OK) using Minesight© software. Estimation was performed using a single pass, as most domains were sampled on a regular grid pattern and samples are thus similarly located with respect to blocks. In addition to grade, the number of informing samples, the distance to the closest sample and the kriging slope of regression were stored also. • Sample selection used during the interpolation process was constrained with a maximum of 3 composites from any given drill-hole allowed to be used to interpolate a block. The minimum and maximum number of composites allowable to interpolate a block was set at 1 and 24 respectively. • The overall Coefficient of Variation for most of the deposits is relatively low (i.e. CV ~ 0.8 – 1.5). Because of the relatively low coefficients of variation observed for available sample composites for each domain area, it further confirmed that OK interpolation method is appropriate. th • The typical upper-cut used is the 99 percentile to contain outliers. • The Ag data is incomplete in some Mineral Resources. Statistical analysis showed that Ag and Au are un-correlated. Silver has been estimated using OK within the gold domain, with unique kriging parameters for each orientation domain. The recovered Ag grade (through mining and milling) shows acceptable reconciliation with the global estimates. Oxide: KHN-Au, TKN-Au, TKM-Au, TKE-Au, PHB-Au

• As a general rule, a gold specific domain was not created for the copper dominant deposits as explained under ‘Geological interpretation’. Rather

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 36 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status there exist certain domains (MnO – PHB & TKN-Au, Limonite – TKM-Au & TKE-Au; Gossan – KHN-Au) that are all near surface oxidation domains weakly coincident with gossan formation. • 2 metre composites were chosen for statistical analysis and grade estimation following investigation of a series of ranges from 1-5m • Statistics were determined by Hellman & Schofield for KHN-Au, by AMC consultants for TKN-Au, TKM-Au, PHB-Au, and by OZ Minerals staff for TKE-Au, using Isatis© software. From this work variography and estimation parameters were developed for copper, gold, silver, molybdenum, sulphur, calculated pyrite, manganese, iron, calcium, magnesium and aluminium. • Gold has a moderate relative nugget effect, ranging from 10-30%. • The majority of the variance occurs within ranges of 10-40m. The longest range structures generally vary between 60-120m. • Elements were interpolated into blocks by OK. - TKE-Au – Block model panel sizes reflected the drill and sample spacing, and domain morphology with parent blocks in the infill drilling grid being 50x50x25m in size (East, North, Rl) and sub- blocking allowed to 5x5x5m in size. 1m drill-hole composites were selected on the basis of summary statistics and were used for variogram analysis and grade estimation. Variogram analysis was undertaken using IsatisTM statistical software. Grades were estimated into hard boundary domains using ordinary kriging and VulcanTM mining software. Variables interpolated include Cu, Au, Ag, Mo, S, Mn and Fe. Search ellipse orientations were determined by geology domain shape and variogram analysis. Each domain was estimated using two passes with increasing search ellipse size (pass 1 - 100:60:30m and pass 2 – 200:120:60m). Minimum number of composites per estimate was 5 with a maximum of 20. Upper cuts were selected on the basis of statistical analysis and are above the 97.5%. - TKN-Au – Two search passes, with the second twice the size of the first. The minimum number of composites used to inform a cell with grade was 4 and the maximum is 10 for the first pass, the second pass minimum was 3, maximum 10. - TKM-Au – For the limonite domain (gold hosting), the Major-Semi- Minor search ellipse is 150-120-60m respectively. The minimum number of samples was 4, with a maximum of 14 with 1 quadrant and opt. per quadrant of 14. A cut-off grade of 2.00 was used (99%). - PHB-Au – For the manganese domain which is the main host of gold, the three domains had 99.5% cut grades of 1.0/1.5/2.0 ppm

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 37 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Au. The minimum / maximum samples in the search parameter used was 4 and 16 respectively, with major-semi-minor ranges of 150/100/25m respectively. - KHN-Au – The deposit was divided into two search ellipsoid orientation zones, reflecting the change in orientation of the supergene blanket, and variography and grade estimation were conducted for each of the mineralised domains in these zones. Three search ellipsoids and sample moisture criteria were used for composite selection and classifying the Mineral Resource estimate; Run 1 – 70x70x6M (E,N,Rl), octant search, min 10 / max 20; Run 2 – 100x100x9m (E,N,Rl), octant search, min 8 / max 30; Run 3 – 120x120x16m (E,N,Rl), non-octant, min 4 / max 30. Primary: DIS, DSC, DSW, NLU

© • Block models constructed in Datamine . • For each deposit, grade was interpolated into blocks with a parent cell size of 25m x 24m x 2.5m (X,Y,Z). Selected on drill spacing. • Models coded to include four different geological and statistical domains used to control estimation of grade into the models (lithology; oxidation; grade shell; mineralisation orientation). • Definition drilling (RC, DD) dominantly sampled at 1m intervals. A 2m composite was chosen after evaluation 1-5m ranges, and validation. © • Geostatistics undertaken by AMC consultants using Isatis . From this, variography parameters were developed for Au, carbon (organic and carbonate), sulphur (total and sulphidic), silver, iron, and arsenic for all deposits. In addition to this, global change of support work to quantify conditional bias issues were undertaken as required. • The average relative nugget effect was around 25-30%. Average drill hole spacing is around 25m, and the longest range structures generally vary between 50-150m. Shortest range structures are generally in the vertical direction (densest composite data). © • Gold grades estimated using OK in Datamine . • Two search passes were required to fill the gold model with grades, the first filled 95% of cells. The second cell was twice as large as the first. A minimum of 4 and maximum of 10 composites were used to inform a cell with grade in the first pass. The second pass used a minimum of 3 and maximum of 10. • The Ag data is estimated as for Au-oxide Mineral Resources. • Drilling has been undertaken to better understand the controls on proportions of sulphide sulphur and organic carbon (critical components for potential processing options) within the mineralisation. The currently understood controls of these elements resulted in the following

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 38 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status domaining: Sulphide Sulphur is domained on oxidation, lithology and gold grade domain, whilst Organic carbon in modelled on oxidation and gold grade domain. Both elements were estimated using OK. Further studies and metallurgical test work continue to refine our knowledge on this. • Grade control (GC) data where available was estimated separately within the gold domain and stored in the model. Modelling • DKY is an ordinary kriged (OK) block model that remains unchanged from techniques 2007. • DIS / DSC are ordinary kriged (OK) block models that remain unchanged from 2007. • DSW is an ordinary kriged (OK) block model and remains unchanged from 2007. • DSE is an update of the 2007 ordinary kriged (OK) block model and includes information from grade control drilling. • LOL is an ordinary kriged (OK) block model and remains unchanged from 2007. • KHN-Au is an update of the 2007 ordinary kriged (OK) block model and includes data from additional grade control, geotechnical and infill drill holes. • NLU is based on depletion of the 2007 ordinary kriged (OK) block model. • NKE is based on depletion of the 2005 ordinary kriged (OK) block model. • NKW is the 2005 ordinary kriged (OK) block model and remains unchanged from 2006. • PHB-Au is an update of the 2007 ordinary kriged (OK) block model created from definition drilling. • PVN is an ordinary kriged (OK) block model and remains unchanged from 2007. • TKE-Au is an ordinary kriged (OK) block model created from historical & exploration drilling during the reporting period. • TKN-Au is an ordinary kriged (OK) block model and remains unchanged from 2007 • TKM-Au is an ordinary kriged (OK) block model, updated from the 2007 MIK model, includes data infill drill holes. • VNE is an ordinary kriged (OK) block model and remains unchanged from 2006. • YNG is an update of the 2007 ordinary kriged (OK) block model from additional infill diamond drilling. Moisture • Moisture has been removed from all tonnage estimates. Cut-off • Visual assessment of the relationship between grade distribution and

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 39 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status parameters underlying geology supports the use of grade-based domains for constraining the estimation of gold Mineral Resources. This is confirmed by statistical analysis which shows a clear step in grade across mineralisation boundaries. • These domains were developed using geological boundaries as a guide to grade envelopes. As a typical example, thresholds of 0.20g/t Au, 0.25g/t Au, 0.30g/tAu and 0.5g/t Au were used for DSW, DIS &DSE-F, YNG & LOL and NLU respectively. Different grade cut-offs were chosen based on apparent grade continuity. • YNG also has a ‘high grade’ domain that is defined by a 4g/tAu cut-off and based on proximity to the main fault and an along strike coherent zone of higher grade. • The copper deposits which also contain gold Mineral Resources below a <0.5%Cu cut-off (TKN-Au; TKE-Au; TKM-Au; KHN-Au) have been defined by a limonitic gossan/clay and a gossan respectively. The controls on gold mineralisation in the PHB copper pit are largely within the MnO domain. Domaining used for estimation was primarily undertaken to constrain the copper estimation. • The final Mineral Resource estimate is then generally reported at 0.5g/t Au, 1.0g/t Au and 1.5g/t Au lower cut-off’s and further tabulated and sub- divided by material oxidation state and Mineral Resource classification confidence. Mining • Currently all mining at Sepon is by open pit method, however, deeper factors or primary Mineral Resources may be more appropriately accessed by assumptions underground mining. • Certain mining factors are applied to Mineral Resource estimation and are detailed below; • The selection of 2m composites took into consideration the typical mineralised thickness, grade statistics and the typical mining bench height being used at Sepon. • In consideration of the data density and mineralised zone geometry factors, the most optimal estimation block size is typically 15m x 6m x 2.5m [East (X); North (Y); Elevation (Z)]. This is an optimal block size that adequately delineates the mineralised zones within the block model, whilst simultaneously not compromising the localised calculated block variances. It is also as close as possible to the Selective Mining Unit (SMU) utilised at the Sepon oxide gold pits. This block size is relatively easy to re-block to 5m x 3m x 2.5m for grade control modelling. Metallurgical • Current gold oxide recoveries are in the range of 75-90%. factors or • The gold Mineral Resources have been defined to separate the three main assumptions oxide zones with respect to metallurgical characterisation. The surfaces are defined as the base of complete oxidation, the base of the transitional

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 40 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status (or mixed oxidation) zone, which also forms the top of the third - primary zone. • Recent drilling has been undertaken to better understand the controls and proportions of sulphide sulphur and organic carbon (the critical elements for potential processing options of primary gold mineralisation). The currently understood controls of these elements resulted in further domaining of some deposits; Sulphide Sulphur is domained on oxidation, lithology and gold grade domain whilst Organic carbon is modelled on oxidation and gold grade domain. Both elements where estimated were Ordinary Kriged (OK). Further studies and extensive metallurgical work are continuing to refine the understanding of these elements and their impact on processing. Bulk density • Samples for bulk density determination are taken from diamond drill core every 10m using a wax immersion method. Twenty centimetre samples are dried for 24 hours in a drying oven. The samples are then weighed as dry samples. They are then coated in wax and weighed in water and then in air. The specific gravity of the wax is also measured to define any differences in specific gravity due to the addition of the wax coating. • The bulk density determinations are the basis for assigning bulk densities to the Mineral Resource estimates and are predominantly based on mineralised/waste zones, consideration of the host lithology and oxidation state. Classification • Mineral Resource classification is based solely on geological and grade confidence. Classification is determined by examination of the following criteria: - Statistical: most commonly kriging variance, occasionally slope of regression is also reviewed. - Sample: factors are often applied based on sample moisture, recovery and type. - Data: the relative data density, distance of nearest composite and number of composites used. - Geological: mineralisation continuity including spatial configuration and spatial continuity. • Classification is applied on a - Block-by-block basis, then - Annealing performed whereby classification solids are constructed around aggregate areas. • Quality indicators are often applied, and the data is reviewed against the overall drill spacing, which can sometimes be used to adjust (often downwards) the classification. • Areas in which RC grade control drilling exists are classified as Measured.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 41 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Additional areas of Measured have sometimes been defined in high confidence Primary areas using the criteria as defined above, however due to the variability experienced in the oxidation surfaces a “buffer” of lower confidence exists through the transition zone.

3.5 Mineral Resource Contributions

The following identifies the key personnel involved in the estimation of the Sepon Gold Mineral Resources:

DIS, DSC, DSW, NLU, NKW, NKE, VNE, PVN and LOL – All drill hole data was sourced and validated by LXML geologists. Geological interpretations and geostatistical analyses were performed either by LXML and Quantitative Geoscience Pty Ltd (QG), AMC consultants Pty Ltd, Ravensgate Pty Ltd (RG) or Geomodelling Pty Ltd (GM) geologists. All significant 2008 updated geostatistical analyses and grade interpolations underwent an external form of review and/or audit with AMC consultants Pty Ltd, or CS-2 Pty Ltd.

Tracie Burrows (AMC). DIS, DSC, DSW and NLU primary gold updates. These estimates were based on the data, geological interpretations and other technical information supplied by LXML.

Mike Stewart (QG). NKE (2005) and NKW (2005). These estimates were based on the data, geological interpretations and other technical information supplied by LXML

Craig Michael (formerly LXML). VNE (2006) - Drill hole data was sourced and validated by and by LXML geologists. Geostatistical analyses, resource modelling and estimation were conducted by Craig Michael under the guidance of Mike Stewart (QG).

Saut Simbolon (formerly LXML). PVN (2007) – Drill hole data was sourced and validated by and by LXML geologists. Geostatistical analysis was conducted by Mike Stewart (QG) in 2006. Resource modelling and estimation were conducted by Saut Simbolon.

Stephen Hyland (RG). LOL (2007) and YNG (2008) – Drill hole data was sourced and validated by LXML geologists, resource domaining was conducted by Stephen Hyland (RG) under the direction of LXML geologists. Block modelling, geostatistical analyses and grade interpolation were conducted by Stephen Hyland (RG).

KHN-Au – Drill hole data was sourced and validated by and under the direction of LXML geologist and Duncan Hackman of Hackman and Associates Pty Ltd (HA). Geostatistical analyses were

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 42 of 132 Mineral Resources 2008 Explanatory Notes conducted by Arnold van der Heyden of Hellman and Schofield Pty Ltd (HS) and resource modelling and estimation conducted by Duncan Hackman (HA).

TKN – Drill hole data was sourced and validated by LXML geologists and Kerrin Allwood (GM) Resource modelling and domaining was conducted by GM under the direction of LXML geologists. Geostatistical analysis was conducted by Mark Sweeny (AMC) and grade interpolation conducted by Tracie Burrows (AMC).

TKM – Drill hole data was sourced and validated by LXML geologists and Kerrin Allwood (GM) under the direction of LXML geologists. Resource modelling and grade interpolation conducted by GM, geostatistical analysis was conducted by Mark Sweeny (AMC), with grade interpolation conducted by LXML geologists, assisted by Steve Hyland (RG).

TKE – Drill hole data and interpretation was conducted by LXML exploration geologists, with geostatistics and estimation completed by Jared Broome (OZ Minerals) in 2008.

PHB - Drill hole data was sourced and validated by LXML geologists. Resource modelling and grade interpolation conducted by LXML geologists with assistance from Steve Hyland (RG), using geostatistical analysis conducted by Mark Sweeny (AMC).

3.6 Competent Person Statement

The information in this report that relates to Mineral Resources is based on information compiled by Jason McNamara and Terry Briggs under the supervision of Paul Quigley a full-time employee of OZ Minerals in the position of Manager – Geology, OZ LXML Sepon, and is a member of the AusIMM and AIG.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 43 of 132 Mineral Resources 2008 Explanatory Notes

Michael Stewart BSc (Geol), MSc (Hons), CFSG (post-graduate Diploma in Geostatistics), MAusIMM, MAIG, employed by Quantitative Geoscience Pty Ltd, has over 20 years experience as a geologist in mining and exploration, which includes 12 years experience in resource estimation and 7 years experience in sediment-hosted gold deposits. He is responsible for geological interpretation and modelling, geostatistical analysis, and grade interpolation.

Jared Broome B.App.Sc (Geol)(Hons), MSc, Grad. Dip. Mining, Grad. Dip. Business, FAusIMM, MAIG, full-time employee of OZ Minerals, has over 16 years experience as a geologist in mining and exploration which includes 13 years experience in estimation of base and precious metal deposits.

Competent Person responsibility is as follows:

DIS, DSC, DSW, NLU, LOL, TKN-Au, TKM-Au, PHB-Au, YNG– Paul Quigley (LXML).

NKW, NKE, VNE and PVN – Paul Quigley (LXML) and Michael Stewart (QG).

KHN-Au – Duncan Hackman (HA).

TKE-Au – Jared Broome (OZ Minerals)

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 44 of 132 Mineral Resources 2008 Explanatory Notes

4 PROMINENT HILL MINERAL RESOURCE STATEMENT - 30 JUNE 2008

4.1 Summary

The Prominent Hill Mineral Resource has increased by 35% (0.65 Mt) to 2.5Mt contained copper and 100% (3.8Moz) to 7.4Moz contained gold due to the conversion of exploration results to Mineral Resources. Additions have come from the Western Au, Eastern, Western and Central- Lower Cu-Au Mineral Resource areas. The total Prominent Hill Mineral Resource has been estimated to be 174.2Mt of copper mineralisation grading 1.39% Cu, 0.56g/t Au and 3.4g/t Ag and 109.2Mt of gold dominant mineralisation grading 0.09% Cu, 1.21g/t Au and 1.0g/t Ag.

Table 6 Prominent Hill Mineral Resource as at 30 June 2008.

Tonnes (Mt) Cu (%) Au (g/t) Ag (g/t) Cu Resource 174.2 1.39 0.6 3 Au Resource 109.2 0.09 1.2 1 Total 283.4 0.89 0.8 2

Table 7 Prominent Hill Copper Mineral Resource as at 30 June 2008

Tonnes (Mt) Cu (%) Au (g/t) Ag (g/t) Measured 44.2 1.71 0.5 4 Indicated 62.7 1.29 0.5 3 Inferred 67.2 1.29 0.6 4

Cu Mineral Resources reported above 0.5% Cu cut-off

Table 8 Prominent Hill Gold Mineral Resource as at 30 June 2008

Tonnes (Mt) Cu (%) Au (g/t) Ag (g/t) Measured 0.4 0.38 0.9 2 Indicated 37.5 0.08 1.1 1 Inferred 71.2 0.09 1.3 1

Au Mineral Resources reported below 0.5% Cu and above 0.5 g/t Au cut-off

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 45 of 132 Mineral Resources 2008 Explanatory Notes

4.2 Setting

The Prominent Hill iron-oxide copper gold (IOCG) deposit is located in the Mount Woods Inlier, in the north-eastern portion of the Archaean to Mesoproterozoic Gawler Craton, South Australia. The Gawler Craton covers approximately 600,000 square kilometres of South Australia. Outcrop is sparse and most of the current understanding of the geology of the Gawler Craton is derived from exploration drilling and geophysical datasets. The Gawler Craton hosts the Olympic Dam, Prominent Hill, Moonta and a number of other smaller and sub-economic copper-gold deposits (e.g. Acropolis, Wirrda Well). Most of these deposits are genetically related to the Gawler Range Volcanic (GRV) – Hiltaba magmatic event which affected the central and eastern portions of the Gawler Craton around 1600-1580Ma. Copper-gold-silver (-U-REE) mineralisation at Prominent Hill is hosted within haematitic breccias of sandstone, shale, and dolomite.

Figure 7 Prominent Hill Project Area, South Australia

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 46 of 132 Mineral Resources 2008 Explanatory Notes

4.3 Changes from the July 2007 Mineral Resource:

The total Cu Mineral Resource for Prominent Hill of 174.2Mt at 1.39% Cu and 0.6g/t Au above a 0.5% Cu cut-off represents a 21.4Mt (14%) increase in tonnes from the July 2007 estimate with increases of 31% in contained copper metal and 34% in contained gold ounces.

The total Au Mineral Resource for Prominent Hill of 109.2Mt at 1.2g/t Au above a 0.5g/t Au cut-off for blocks less than 0.5% Cu represents an increase of 71Mt (186%) from the July 2007 estimate with an increase of 212% of contained gold ounces.

The Open Pit Cu-Au Mineral Resource remains largely similar to the July 2007 Mineral Resource estimate with only minor changes due to triangulation and block model construction.

The total Mineral Resource within the Central Mid Cu-Au area remained comparable to the July 2007 Mineral Resource. Definition drilling conducted since July 2007 has converted 8.6Mt of Cu Mineral Resources and 12.3Mt of Au Mineral Resources from Inferred to Indicated Mineral Resources.

The Central Lower Mineral Resource area has increased by 5.5Mt (Cu Mineral Resource) and 28.9Mt (Au Mineral Resource) compared with the July 2007 Mineral Resource estimate, which is a result of exploration drilling and geological interpretation. In addition, definition drilling has upgraded 7.7Mt of Cu Mineral Resources and 8.5Mt of Au Mineral Resources from Inferred to Indicated Mineral Resources.

The Western and Eastern Cu-Au Mineral Resource areas represent newly added Inferred Mineral Resources as part of the June 2008 Mineral Resource estimate, which are the result of exploration drilling along strike of the Prominent Hill Open Pit Mineral Resource. The Western Au Mineral Resource area located directly west of the Prominent Hill Open Pit also represents newly added Inferred Mineral Resources for June 2008. The Western Au Mineral Resource area is defined by pre-2008 exploration drilling that has recently been interpreted and estimated for consideration in underground mining scoping studies.

Indicated and Inferred categories for both the Cu and Au Mineral Resources increased significantly in tonnes and metal, which is largely due to exploration and definition drilling undertaken between July 2007 and June 2008. The Indicated Mineral Resource tonnage increase is 16.3Mt (35.2%) for the Cu Mineral Resource and 20.8Mt (123.5%) for the Au Mineral Resource. The Inferred Mineral Resource tonnage increase is 3.7Mt (5.8%) for the Cu Mineral Resource and 49.8Mt (233%) for the Au Mineral Resource.

Measured category of the Cu and Au Mineral Resources tonnes and metal increased by small margins of 1.3Mt (Cu Mineral Resource) and 0.4Mt (Au Mineral Resource).

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 47 of 132 Mineral Resources 2008 Explanatory Notes

Figure 8 Long projection of Prominent Hill showing the outline of the 2007 and 2008 Mineral Resources

4.4 Assessment and Reporting Criteria

The following table provides a summary of important criteria related to the assessment and reporting of the Prominent Hill Copper/Gold Mineral Resource.

Criteria Status Sampling Techniques and Data

Drilling • Diamond (NQ2) standard tube drilling and reverse circulation (RC) was techniques used for geological interpretation.

Drill sample • Core recovery was good with 93% recovered. recovery TM Logging • Core was logged into the OCRIS logging system. • Core was photographed prior to being logged by the geologist. • All core is stored at the Prominent Hill core shed.

Sub-sampling • Core is orientated along the bottom of hole and then half-core samples techniques and are taken using a diamond core saw. sample • Sample interval density was measured using “Archimedes Principle” preparation • Samples were dried, crushed and pulverised to a nominal 90% passing - 75 microns.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 48 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Quality of assay • The 2004-08 Cu grades were determined by modified aqua- data and regia/perchloric acid digest ICPOES determination (AMDEL Adelaide ore- laboratory tests grade Cu method) • The 2004-08 Au grades were determined by 40g Fire Assay AAS (at AMDEL Adelaide). • For the pre 2004 data, the AMDEL ore-grade Cu method was adopted for assays greater than 1%Cu in the initial HF/mixed-acid digest ICPOES assay results. • Assay data quality was determined through submission of field and laboratory standards, blanks and repeats which were inserted at a nominal rate of 1 each per 25 drill samples.

Verification of • Umpire laboratory checks were completed during the Bankable Feasibility sampling and Study (BFS) and no issues were identified that would prevent the assaying unreserved classification of the Cu and Au Mineral Resources under the guidelines of the JORC Code (2004 Edition). • Three pairs of twinned holes were drilled at Prominent Hill and their results are detailed in the BFS. TM Location of • All diamond drill holes were surveyed and recorded in the PH-MST GBIS Data points database. • Most drill-hole collar coordinates were surveyed in MGA94_53 using differential GPS (DGPS), however holes drilled prior to 2006 were surveyed using tape and compass or GPS. 12 new drill holes were awaiting DGPS surveys at the time of modelling. These lay on the margins of the deposit and their proposed collars are only expected to change by +/- 1m once surveyed. • MGA coordinates were converted to local mine grid for Mineral Resource estimation. • All drill-holes have magnetic down-hole surveys taken at 30m intervals using either a single or multi-shot down-hole camera. An azimuth adjustment of +6.3o degrees was applied for the conversion to local mine grid for all magnetic surveys. • Some diamond holes were partially surveyed by Surtron or Northern Exploration Services using a North Seeking Gyro (NSG). • During the BFS, an assessment of 30 drill hole traces defined by both North Seeking Gyro and reliable multi-shot data (determined from the magnetic field intensity and declination data collected with each reading), showed separations of less than 10m at the end of hole locations. The NSG data also showed that, apart from the Cadna-owie sand unit in the cover sequence, holes are generally predictable in the degree and direction of deviation. • Obtaining complete survey data for partially gyro-surveyed holes, involved adjusting multi shot magnetic data and adding this onto the last

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 49 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status down-hole gyro survey point.

Data spacing • Drilling has been completed on nominal north-south 50m sections with and 25m infill holes and sections located immediately below the unconformity distribution and in the Eastern Gold dominant zone (also tested with an additional oblique drilling grid). • Within the upper copper-gold mineralisation, holes are drilled approximately 50m apart on section; the eastern gold dominant mineralisation has been drilled at approximately 25m horizontal separation on section. • Within the deeper copper-gold and Western Gold mineralisation, holes are drilled approx. 100-200m apart on section. • The majority of upper holes are angled at approximately 60 degrees to the south, whilst deeper holes are angled at approximately 60 degrees to the north. • There are 7 vertical holes and 29 holes drilled oblique to the drill sections • Drilling is predominantly concentrated between 54000E and 56700E and between 10210RL and 8700RL

Orientation of • The majority of drilling has been completed on nominal north-south data in relation sections which intersect the strike of mineralisation. to geological • There is no expected bias due to the continuity of mineralisation along structure strike. • The intersection angle is between 30 and 50 degrees through the haematite breccia bodies. Estimating and Reporting of Mineral Resources

Database • The Prominent Hill database is a SQL system and consists of three integrity components. These are a Field Logging System (FLS), a transfer-database (TRN), and a master-database (MST). TM • The FLS consists of OCRIS logging software collecting data directly to SQL Server. The software controls data input via pick lists ensuring adherence to logging legends along with their checks. • SQL stored procedures are employed to migrate data from individual logging-computers into the TRN. • The TRN works as a quarantine and compilation system with suitable checking procedures applied. • Data is transferred from the TRN to the MST via SQL Server stored procedures once data has been checked. • Validation checks are written into the SQL database and these are activated via database and user triggers to ensure the MST data is correct with respect to fundamental quality issues.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 50 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Geological • The mineralisation at Prominent Hill forms part of a large regional interpretation alteration system. Interpretation and geochronological analysis of drill samples from Prominent Hill and surrounding prospects suggests a genesis related to the Gawler Range Volcanic / Hiltaba volcano-plutonic event (ca. 1585Ma). • The mineralised system is built on regional E-W, NW and NE-trending mineralizing structural channels which carried copper and gold bearing hydrothermal fluids. Copper-gold mineralisation is hosted by haematite- matrix breccias, which have undergone extensive near-surface sericite alteration and silica hydrothermal overprinting (hydrolytic alteration). • The geology at Prominent Hill is categorized into the following lithologies and structural zones from north to south: • Skarn-granitoid package: a deeper, older package of deformed metasediment, intruded by high-level granitoids and intensely metasomatised during the mineralizing event. • Hanging-wall fault zone (HWFZ); representing the northern extent of the main copper and gold mineralisation, containing chloritic fault breccias, graphitic shear zones, skarn, granitoid and dolomite in an east-west trending zone of 10-20 metres true thickness. It separates the skarn- granitoid package from the volcano-sedimentary sequence and is the fault set along which juxtaposition of the two, through reverse fault (N- block up) movement, occurred. • Volcano-sedimentary package – Variably haematised and silicified dolomite and limestone, shale, sandstone, greywacke and mafic to intermediate volcanics. • Copper-gold mineralisation occurs in domains of haematite-matrix breccia within the volcano-sedimentary package. • Several late stage dolerite dykes cross-cut the mineralisation and are thought to represent Gairdner Dyke Swarm equivalents (ca. 800Ma).

Dimensions • Mineralisation zones are tabular and sub-parallel to the lithostratigraphic architecture in the west and are pipe-like and plunge steeply WNW to NW in the east. The deposit is thought to have formed in a north-dipping reverse fault system between ENE and WNW trending faults and shear zones, with breccias forming in zones of maximum dilation. • The dip of the mineralisation is sub-vertical to steep northerly. • Known zones of mineralisation extend from 54000E to 56700E, 11800N to 12800N and 10210RL and 8700RL

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 51 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Estimation and • Polygons and hence triangulations are based on interpretations Modelling completed on 50m easting sections. Techniques • Triangulated interpretations have been domained into the following constrained fault and breccia bodies: - HWFZ - PHSZ - BD1 - BD2 - BD3 - BD4 - BG1 - BG2 - BG3 - FW-Au - HW-Au - Northern Cu - Western Cu, and - Eastern Cu • In addition to these mineralised breccia domains, lithological domains, (±Cu/Au mineralisation), have also been constructed. These include: greywacke-sandstone (GWK-SST), DOLOMITE, SHALE, SKARN, MSED, quartz-microdiorite (QMDT) and andesite (ANDES-1, -2 and -3). • Late stage, unmineralised dolerite dykes have been domained as DOLT-1, -2 and -3. • Metallurgical characteristics (derived from early test work conducted in 2004) are incorporated into the Mineral Resource domaining and provide further controls on grade interpolation. The metallurgical/mineralisation domains within the breccias are modelled separately as either chalcocite- bornite (CCBN), pyrite-bornite (PYBN), bornite-chalcopyrite (BNCP) or chalcopyrite-pyrite (CP). The mineralogical divisions are based on visual estimates from logging, Cu:S ratios and Co assays (as pyrite at Prominent Hill contains elevated Co). • A priority system of 48 domains was set up to account for overlapping mineralisation, intrusive rock shapes and cover sequence lithologies. • Priority numbers ranged from 50 (lowest priority), to 910 (highest priority). A 49th code of 950 was assigned to all blocks above the surface within air. • The block model was constructed with parent blocks of 25mE by 25mN by 12mRL within areas with average drill spacing up to 50m by 50m, other areas applied a parent block size of 100mE by 100mN by 100mRL. • Ordinary kriging (OK) to the parent block size was used to estimate Cu,

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 52 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Au, Ag, U, Fe, Ba, S, Si, Ca and F grades separately. • Up to three estimation passes with increasing search neighbourhood size were run for all domains (Pass1 – 75 by 75 by 10m, Pass 2- 150 by 150 by 10m, Pass 3 – 300 by 300 by 20m). • A minimum of 4 and maximum of 32 composites were used per estimate. • An Octant based search limited composites to a maximum of 4 composites per octant (not applied to Pass 3 for domain 610). • 5m assay composites were used. • Estimation applied composite length weighting.

Moisture • Tonnes have been calculated on a dry basis.

Cut-off • Copper Mineral Resources have been reported above a 0.5% Cu block parameters grade cut-off, and • Gold dominant Mineral Resources have been reported below a 0.5% Cu and above a 0.5 g/t Au block grade cut-off. • The reporting cut off was increased from the July 2007 Mineral Resource (0.3% Cu cut-off) to improve reporting of the global Mineral Resource, which currently has a large component of mineralisation that may be extractable by underground mining methods.

Mining factors • The upper-central portion of the Mineral Resource is currently being or assumptions mined as an open pit operation. • Underground mining studies are being undertaken on Mineral Resources outside the planned open pit area.

Metallurgical • Metallurgical characteristics within the total copper and gold Mineral factors or Resources respectively are considered to be comparable with areas assumptions studied as part of the BFS.

Bulk density • All sampled core and more than 20% of all other core has been measured for density. • The method used the entire air-dried core sample weighed in air and water, which was used to estimate the density. • Regression analysis of iron assays and density was applied to estimate the density of blocks given the ordinary kriged iron value. Where blocks were not estimated for iron, the average density for the domain was assigned. • Several waste domains with low iron values were assigned the average domain density.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 53 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Classification • Mineral Resources have been classified on the basis of geological confidence. • The method of classification of Open Pit Mineral Resources remains unchanged from the 2006 estimate, which applied resource categories on the basis of mineralised domain combined with drill-hole spacing. All other Mineral Resource areas were classified as follows: • Measured Mineral Resources have an average drill-hole spacing of 25 by 25 metres. • Indicated Mineral Resources have an average drill-hole spacing of 50 by 50 metres, and • Inferred Mineral Resources have an average spacing of up to 100 by 200 metres in all areas except the Western Au area which has a drill-hole spacing of up to 200 by 200 metres.

Audits or • A review of this Mineral Resource estimate has been conducted by Behre Reviews Dolbear Australia. • An audit and review of sampling techniques and data has been undertaken for Prominent Hill during the BFS. • Annual database and protocol reviews are conducted by Duncan Hackman of Hackman & Associates. A detailed audit history pre BFS can be seen in the BFS document.

4.5 Competent Person Statement

Patrick Say B.Sc (Geol)(Hons), MAusIMM, full-time employee of OZ Minerals has over 8 years experience as a geologist in mining and exploration.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 54 of 132 Mineral Resources 2008 Explanatory Notes

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 55 of 132 Mineral Resources 2008 Explanatory Notes

5 CENTURY MINERAL RESOURCES STATEMENT - 30 JUNE 2008

5.1 Summary

The Mineral Resource at the OZ Century Mine as at 30 of June 2008 is summarised in Table 9 and Table 10. The Eastern Fault Block estimate remains unchanged since the 31st of March 2008 Mineral Resource.

Table 9 Century Mine Mineral Resource as at 30 June 2008

Tonnes Ag Block Zn (%) Pb (%) (Mt) (g/t) Measured 36.6 12.3 1.4 33.7 Indicated 10.8 11.7 1.4 35.5 Inferred - - - - Total 47.4 12.2 1.4 34.1

Table 10 Eastern Fault Block Mineral Resource as at 30 of June 2008

Tonnes Ag Block Zn (%) Pb (%) (Mt) (g/t) Measured - - - - Indicated 0.2 12.8 1.1 49.0 Inferred 0.2 12.7 1.1 55.0 Total 0.4 12.8 1.1 52.0

The Mineral Resource is reported at a 3.5%Zn cut-off.

5.2 Introduction

The OZ Century Mine is a Zinc-Lead-Silver deposit located approximately 250km North-West of Mount Isa in Far North Queensland. The mineralisation is hosted in Proterozoic shales interbedded with siltstones capped with barren sandstone and Cambrian limestone. The eastern fault block mineralisation occurs within a Proterozoic block that has been faulted into the younger Cambrian limestone sequence.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 56 of 132 Mineral Resources 2008 Explanatory Notes

The OZ Century Mineral Resource is based on the April 2008 Mineral Resource block model (cenapr2008.bmf) and the April 2008 Eastern Fault block model (cenefb2008.bmf). These models were completed by Snowden Mining Industry Consultants (Snowden) and are based on seam accumulation from a reference surface to generate the volume model and ordinary kriging of full seam composites for the estimation of grades. No changes were made to the block model during the previous 3 months and all changes to the current Mineral Resource were due to mining depletions. The Mineral Resource has been classified according to the guidelines of the JORC (2004) code and considers drill hole spacing, estimation results and the internal and the bounding structures of the deposit.

The Mineral Resource estimate is based on drilling conducted between 1990 and 2007. The geological and structural interpretation was carried out by OZ Minerals mine geologists, and the block model and grade estimation were completed by Snowden using VulcanTM software.

5.3 Setting

The Century Zinc-Lead-Silver mine is located in the remote lower Gulf region of north-west Queensland, approximately 250km north-west of Mount Isa.

Figure 9 Century Mine Location

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 57 of 132 Mineral Resources 2008 Explanatory Notes

The deposit is hosted within the Lawn Hill Formation, a Middle Proterozoic sequence of shale, siltstone and sandstone overlain by younger Cambrian limestone. Structurally, the deposit is located within the Page Creek syncline and is terminated to the east by Cambrian limestone and faults associated with the Termite Range Fault. Magazine Hill Fault and Nikki’s Fault define the southern and northern boundaries respectively. The western boundary is truncated by Cambrian limestone and by present day surface at the Discovery Hill gossan. The mineralisation is divided into northern and southern blocks by the north-dipping normal Pandora’s Fault.

High-grade mineralisation at Century occurs mostly in black shales, dominantly as fine grained sphalerite and galena lamellae with siderite and minor pyrite. The black shale units are separated by less mineralised, siderite rich, siltstone horizons. The deposit is unmetamorphosed, only weakly deformed, and displays excellent lateral stratigraphic and grade continuity apart from small-scale fault dislocations.

5.4 Assessment and Reporting Criteria

The following table contains a summary of important criteria and data used in the estimation and assessment on the Century Mineral Resource.

Criteria Status Sampling Techniques and Data Drilling • The geological interpretation is based on over 1000 drill holes, of which techniques 421 contained valid intersections and were used in grade estimation; • The drilling types used in the Mineral Resource estimate include NQ and HQ size diamond drill core. Drill Sample • Drill core recovery within the mineralised sequence is typically 100 Recovery percent; • All diamond drill holes have had core recovery recorded, along with other basic geotechnical information. Logging • Detailed geological logging was completed on all drill holes using the refined stratigraphy developed by Solid Geology (2002). Paper logs were then entered in an access database; • A geophysical logging system is used to log production blast holes and has also been used to log deeper RC holes. This system measures natural gamma radiation and magnetic susceptibility to differentiate units within the mineralised zone.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 58 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Sub-sampling • Half core samples are taken using a diamond core saw; techniques and sample • Sample intervals are confined to geological boundaries, determined by preparation the geologist responsible for logging the hole, with depth and unit information recorded;

• In the 1990-95 drilling campaigns;

- Sample preparation of all drill samples were completed by Amdel at their Mt Isa facility;

- Samples of approximately 4 to 5 kg were dried in aluminium trays;

- Samples were jaw crushed to 6mm;

- Samples were pulverised to 70% passing 75 micron in a mixermill with three concentric rings;

- A split of 200g was bagged and sent for assay;

- The sample residue was re-bagged and stored at the CRAE facility in Canberra;

• In the 1999 – 2007 drilling campaigns;

- Sample preparation of all drill samples were completed at the Century Mine laboratory;

- Samples were dried and the entire sample crushed to 5mm through a jaw crusher;

- Samples were riffle split to produce 300gm for pulverising;

- Sample pulverised to minus 200 microns in a ring mill.

Quality of assay • Samples from 1990-95 were assayed through Amdel Laboratories in data and Mount Isa and Adelaide, Analabs in Townsville and Genalysis Laboratory laboratory tests Services, Perth. Analytical methods used were Atomic Absorption Spectrometry (AAS), Induced Coupled Plasma Optical Emission Spectroscopy (ICP-OES) and Leco furnace methods. • Post 1999 samples are analysed in the Century mine laboratory using a combination of x-ray fluorescence (XRF – Zn, Pb, Fe, Mn, SiO2, S), atomic absorption spectrometry (AAS – Ag) and Leco furnace (total carbon) methods, and each month a sub-set of samples are despatched to ALS in Brisbane for check assay. Each internal and external batch of samples is accompanied by selected Standard reference sample material.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 59 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Verification of • The QAQC controls for both sets of drilling campaigns included; sampling and - The insertion of a variety of laboratory certified standard samples assaying based on Century mineralisation; - Duplicate samples of quarter core; - Duplicate samples of 5mm splits (Century lab only); - Submission of pulps to off site “umpire” laboratories - Repeats of assayed pulps. Location of data • Collar co-ordinates of all drill holes were determined to an accuracy of points 0.1m in all directions by a licensed surveyor. • Down-hole surveys were taken at 30m intervals for all inclined holes and 30 – 40 percent of vertical holes using single shot Eastman cameral equipment Data spacing and • Drill-hole collars are located on an approximate grid pattern with a distribution spacing of between 50 and 70 metres on north-south sections across the deposit.

Orientation of • The Century mineralised sequence dips at between 5 and 25 degrees data in relation to over most of the deposit area, with dips up to 70 degrees around the geological margins. structure • The majority of drill holes are therefore vertical with inclined holes targeted at the more steeply dipping zones.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 60 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Audits or reviews • In 1996 Mining and Resource Technologies (MRT) completed a data validation and review of the initial drilling completed by CZL from 1990 – 1995 (MRT0378 Feb 1996 and MRT0435 Jul 1996) • In 2002 and 2003 Snowden completed reviews on the data quality and QAQC procedures for geology sample data from 1999 – 2003. Estimating and Reporting of Mineral Resources Database • All drilling, sampling, assay, density and geological data is stored in the integrity Microsoft Access database PCMGeolData.mdb • The geology database has been validated and audited by independent parties Snowden and MRT • All data is entered manually into access database forms with look up tables to ensure data validation Geological • Mineralisation in the deposit is strata bound and confined by interpretation stratigraphy, which is well understood and well defined throughout the deposit. • The interpretation of the deposit geology was based on all available drilling information at the time of model generation • In addition to the information gained from geological logging, down- hole dip-meter information was used to verify structural interpretations and geophysical probing of blast holes in the mineralised sequence are used in the geological interpretation along with mapping data collected during mining • There are a number of bounding structures which limit the model extents: - Magazine Hill Fault forms the southern boundary of the deposit, which has an east-west strike, dips to the north and has a North Block down offset. - Nikkis Fault forms the northern boundary of the deposit and is the northern wall of the graben in which the deposit lies. The fault strikes east-west and dips steeply to the south. - Pandoras Fault is the major boundary between the North and South Blocks. Vertical offsets range from 5 m at the eastern margin to greater than 200 m at the western edge of deposit. - The overlying Cambrian Limestone forms the boundary to the deposit along the eastern margin of the South and North Blocks of the deposit. It also forms the boundary to the western margin of the North Block. • There are also several modelled internal structures that displace mineralisation by various amounts.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 61 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Dimensions • The Century mineralisation extends from 26,850N to 28,350N, 46,400E to 47900E and 1125RL to 814RL. Estimation and • Surfaces were modelled for each of the stratigraphic units in the modelling mineralised sequence techniques • Three types of composites were generated as part of the Mineral Resource model construction - Vertical thickness composites - Grade composites (Zn, Pb, Ag, Fe, Mn and S) - Density composites • Compositing was done to generate single samples over the full length of each unit at each drill-hole intercept. • The grade statistics of each unit were reviewed. The sample populations were reviewed for each Unit to identify the type of distribution and identify any outliers or characteristics that would have implications for variogram modelling and grade estimation. • Top cuts were applied where appropriate: Unit/Element Topcut Number of samples cut 150 zinc 18.0% Zn 1 185 zinc 10.0% Zn 2 150 lead 6.0% Pb 6 170 lead 30.0% Pb 1 175 lead 6.5% Pb 3 185 silver 400 g/tAg 2 200 silver 400 g/tAg 1 • Traditional variogram interpretation using spherical models was undertaken on elements zinc, lead, silver, iron, manganese and sulphur for each unit in the mineralised sequence and also the vertical thickness and density variable for each unit in order to determine suitable search parameters. • The volume model was constructed by combining the estimated vertical thickness of each Unit with a detailed model of the reference surface and bounding structures. It followed the seam accumulation method used in the April 2005 Mineral Resource model and attempts to honour the stratigraphic nature of the deposit. The reference surface used was the base of Unit 320. This surface was chosen as it is visually distinct, has sharp geophysical characteristics and is near the centre of the deposit, minimising any errors due to accumulation of estimated vertical thickness

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 62 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status • A kriging neighbourhood analysis study was undertaken on the estimation parameters for units 200 and 410 (the main metal bearing units) in order to minimise any conditional bias. This helped to establish preferred block sizes and the number of samples used to estimate values in each block • Block model origin and extents: Dimension Minimum Maximum Parent Number Sub Cell (m) (m) Cell Size of Cells Size (m) (m) Easting 45900 48300 10 240 10 Northing 26300 28700 10 240 10 Elevation 800 1200 400 1 0.05 • Ordinary Kriging was used to estimate zinc, lead, silver, iron, manganese, sulphur, thickness and density for each unit.

Moisture • Tonnes have been calculated on a dry basis

Cut-off • The Mineral Resource is reported at a 3.5%Zn cutoff parameters • No assumptions were made regarding cut-off grade for the Mineral Resource due to the deposit being restricted to certain strata, which are themselves constrained by structural surfaces and topography

Mining factors • No assumptions were made with regard to mining factors in the Mineral and assumptions Resource model.

Metallurgical • No assumptions were made with regard to metallurgical factors in the factors or Mineral Resource model. assumptions

Bulk density • The process of estimating bulk density involves calculating the stoichiometric density of composites, applying a correction factor for porosity based on grab sample results and then estimating the corrected stoichiometric density using Ordinary Kriging. - Select samples that have assay results for all elements required in the stoichiometric equation. - Composite these samples for intervals that have been coded as having valid samples. - Calculate the stoichiometric density for these samples - Apply correction factor derived from the grab sample bulk density to convert the stoichiometric density into bulk density - Derive variograms for bulk density for each Unit - Estimate bulk density into block model.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 63 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Classification • The Mineral Resource has been classified according to the guidelines of the JORC code (2004) and takes into account the drill hole spacing, estimation results and the internal and bounding structures of the deposit. The model variable class has been coded as either measured (class = 1) or indicated (class = 2) • Mineral Resources have been classified as measured where the mineralised sequence has been interpolated rather than extrapolated from drill hole data and is greater than 5m from interpreted major internal structures • The Mineral Resource outside the polygon that defines the boundary of composites for the main Units where grades have been extrapolated rather than interpolated is classified as Indicated Mineral Resource. In this area there is less confidence in the estimated grades as shown by the decreased kriging efficiency (KE) and regression slope statistics, and also the lower confidence in the deposit extent due to uncertainties in the position of the bounding structures. Blocks adjacent to internal structure have also been classified as indicated to reflect the lower confidence in spatial location of the mineralised Units due to the offset associated with these structures. A 5m buffer around the trace of these faults was used to flag blocks as indicated Audits or reviews • The Century Mineral Resource calculation was carried out by Snowden and reviewed by OZ Minerals staff.

The following table contains a summary of important criteria and data used in the estimation and assessment on the Century Eastern Fault Block Mineral Resource.

Drilling • The geological interpretation is based on 8 diamond drill holes that techniques intersected the mineralised zone

Drill Sample • Drill core recovery within the mineralised sequence is typically 100%. Recovery • All diamond drill holes have had core recovery recorded, along with other basic geotechnical information.

Logging • Detailed geological logging was completed on all drill holes and paper logs were later entered into an access database.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 64 of 132 Mineral Resources 2008 Explanatory Notes

Sub-sampling • Half core samples are taken using a diamond core saw techniques and • Sample intervals are confined to geological boundaries, determined by sample the geologist responsible for logging the hole, with depth and unit preparation information recorded. • In the 1990-95 drilling campaigns; - Sample preparation of all drill samples were completed by Amdel at their Mt Isa facilty - Samples of approximately 4 to 5 kg were dried in aluminium trays - Samples were jaw crushed to 6mm. - Samples were pulverised to 70% passing 75 micron in a mixermill with three concentric rings. - A split of 200g was bagged and sent for assay - The sample residue was re-bagged and stored at the CRAE facility in Canberra. • In the 1999 – 2007 drilling campaigns; - Sample preparation of all drill samples were completed at the Century Mine laboratory. - Samples were dried and the entire sample crushed to 5mm through a jaw crusher. - Samples were riffle split to produce 300gm for pulverising. - Sample pulverised to minus 200 microns in a ring mill.

Verification of • The QAQC controls for both sets of drilling campaigns included: sampling and - The insertion of a variety of laboratory certified standard samples assaying based on Century mineralisation; - Duplicate samples of quarter core; - Duplicate samples of 5mm splits (Century lab only); - Submission of pulps to off site “umpire” labs; - Repeats of assayed pulps.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 65 of 132 Mineral Resources 2008 Explanatory Notes

Location of data • Collar co-ordinates of all drill holes were determined to an accuracy of points 0.1m in all directions by a licensed surveyor. • Down hole surveys were taken at 30m intervals for all inclined holes and 30 – 40% of vertical holes using single shot Eastman cameral equipment.

Data spacing and • The Eastern Fault Block is defined by 8 drill holes. The drill spacing varies distribution from 25m to 50m.

Orientation of • The Eastern Fault Block mineralised sequence is much steeper dipping data in relation to that most of the Century deposit. For this reason most of the holes were geological drilled at an incline to get close to true width intersections. structure

Reviews or audits • In 1996 Mining and Resource Technologies (MRT) completed a data validation and review of the initial drilling completed by CZL from 1990 – 1995 (MRT0378 Feb 1996 and MRT0435 Jul 1996). • In 2002 and 2003 Snowden completed reviews on the data quality and QAQC procedures for geology sample data from 1999 – 2003. Estimating and Reporting of Mineral Resources

Database • All drilling, sampling, assay, density and geological data is stored in the integrity Microsoft Access database PCMGeolData.mdb. • The geology database has been validated and audited by independent parties Snowden and MRT. • All data is entered manually into access database forms with look up tables to ensure data validation.

Geological • The mineralised Units within the Eastern Fault Block have similar interpretation characteristics to the main deposit and the same sequence of Units can be applied. The mineralised sequence dips at approximately 65° towards the north northwest (345°). The mineralised Units are bounded by faults along the lower and southern edges and haematite alteration associated with the paleo-weathering surface removes mineralisation from the upper section of the fault block. The bounding structures were based on the interpretation by Solid Geology (2002) and updated with mapping data from the outcrop exposed in the east wall of stage 2 and additional drill hole ZCL563. • The geological interpretation was completed on a series of 25 m spaced sections, across the strike of the Eastern Fault Block. Solid wireframe models were constructed from the outline of individual Units.

Dimensions • The steeply dipping Eastern Fault Block extends from 48043N, where it outcrops in the east wall of the pit to 48223N, 27130E to 27223E and 1129RL to 1055RL.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 66 of 132 Mineral Resources 2008 Explanatory Notes

Estimation and • Samples were composited by geological Unit to produce single full seam modelling composites for each Unit. techniques • The volume model was rotated to allow sub blocking to best represent the steeply dipping orientation of the mineralised sequence. The orientation of the model is in the plane of mineralised sequence and dips at 65 degrees to the north northwest. The parent block size and sub- celling was the same as used in the main deposit Block model origin and extents:

Dimension Origin Extent Parent Number Sub Cell Cell Size of Cells Size (m) (m) Easting 48000 600 10 60 10 Northing 26900 500 10 50 10 Elevation 1200 400 400 1 0.05

• Inverse distance squared was selected to estimate the grades within the Unit of the eastern fault block. As with the main deposit each Unit was treated as a separate domain and only composites from within that domain were used in the grade estimation. The search ellipse was orientated sub parallel to the mineralised sequence and all blocks were estimated in the first pass Moisture • Tonnes have been calculated on a dry basis.

Cut-off • The Mineral Resource is reported at a 3.5%Zn cutoff. parameters

Mining factors or • No mining assumptions have been made. assumptions

Metallurgy • No assumptions were made with regard to metallurgical factors in the factors or Mineral Resource model. assumptions

Bulk Density • Due to the low number of composites containing the complete set of elements required for the stoichiometric density calculation, the bulk density for each Unit was assigned the average density of each related Unit within the main deposit

Classification • The Eastern Fault Block mineralisation was classified on the drill spacing and whether the model had been interpolated or extrapolate • Due to the additional structural complexities of the Eastern Fault Block compared to the main deposit the Mineral Resource has been classified as Indicated and Inferred. The area interpolated between the drill holes has been classified as indicated and the area extrapolated along strike to the east and west as inferred

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 67 of 132 Mineral Resources 2008 Explanatory Notes

Audits or reviews • The Eastern Fault Block Mineral Resource calculation was carried out by Snowden and reviewed by OZ Minerals staff.

5.5 PRODUCTION DEPLETION

The Mineral Resource at 30 June 2008 was developed by depleting the 31 March 2008 (Table 12) by 3 months to the 30 June 2008. The Mineral Resource depletion for the 3 months ending 30 June 2008 at a cut-off grade of 3.5% Zinc is summarised in Table 11. Mineral Resource depletion was completed by mining out blocks from the April 2008 Mineral Resource block model (cenapr2008.bmf) using pit shells generated from survey pick-up (provided by OZ Century Survey) and Vulcan software. The depleted Mineral Resource reported was contained within the end of June 2008 pit shell after the areas mined up to the end of March 2008 were depleted from the April 2008 model. The Mineral Resource that remained underneath the pit shell in areas that were completely mined out during the 12 month period was added to the total depletion.

The forecasted Mineral Resource remaining (Table 13) for 30 June 2008 compares closely to the Mineral Resource remaining estimated at the 30th June 2008 (Table 14).

Table 11 Mineral Resource depletion for the 3 Months to 30 June 2008

Tonnes (Mt) Zn (%) Pb (%) Ag (g/t) Measured 1.13 14.7 2.0 23 Indicated 0.03 12.6 2.1 24 TOTAL 1.16 14.6 2.0 24

Table 12 Mineral Resource as at 31 March 2008

Tonnes (Mt) Zn (%) Pb (%) Ag (g/t) Measured 37.7 12.4 1.4 33 Indicated 10.9 11.7 1.4 36 Inferred - - - - Total 48.6 12.3 1.4 34

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 68 of 132 Mineral Resources 2008 Explanatory Notes

Table 13 Forecast Mineral Resource as at 30 June 2008

Tonnes (Mt) Zn (%) Pb (%) Ag (g/t) Mineral Resource March 2008 48.6 12.3 1.4 34 3 Month depletion 1.2 14.6 2.0 24 Forecast Resource March 2008 47.4 12.2 1.4 34

Table 14 Comparison of the Forecast Mineral Resource against the actual Mineral Resource at 30 June 2008

Tonnes (Mt) Zn (%) Pb (%) Ag (g/t) Forecast Resource Remaining 47.4 12.2 1.4 34 Reported Resource Remaining 47.4 12.2 1.4 34

Difference 0.1% 0.4% 3.34% 0.5%

5.6 Competent Person Statement

The information in this report that relates to Mineral Resources is based on information compiled by Mr Stefan Mujdrica, who is a Members of the Australasian Institute of Mining and Metallurgy and has sufficient experience which is relevant to the style of mineralisation and type of deposit under consideration and to the activity which they are undertaking to qualify as Competent Persons as defined in the 2004 Edition of the ‘Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves’. Mr Stefan Mujdrica consents to the inclusion in the report of the matters based on their information in the form and context in which it appears.

Mr Stefan Mujdrica is employed by Snowden Mining Industry Consultants.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 69 of 132 Mineral Resources 2008 Explanatory Notes

6 DUGALD RIVER MINERAL RESOURCE STATEMENT - 30 JUNE 2008

6.1 Summary

Dugald River Mineral Resources as at June 30th 2008 are estimated to be 53.8Mt grading 12.7% Zn, 2.0% Pb and 39g/t Ag above a 6% Zn cut off. The breakdown of Mineral Resources by classification category is shown in Table 15

Table 15 Dugald River Mineral Resource as at 30 June 2008 (6% Zn cut off)

Category Tonnes M Zn (%) Pb (%) Ag (g/t) Fe (%) Mn (%) Measured 20.7 13.3 2.0 56 11.7 0.71 Indicated 24.2 12.6 2.0 33 11.6 0.88 Sub total 44.9 12.9 2.0 43 11.7 0.80 Inferred 8.9 11.6 1.8 15 11.2 0.93 Total 53.8 12.7 2.0 39 11.6 0.82

The updated Mineral Resource represents a 12.3% (5.9Mt) increase in tonnes along with a 17.9% increase in contained zinc metal and a 7% increase in contained lead metal. In addition, 20.7Mt of Mineral Resources have been upgraded to Measured and approximately 13Mt of Inferred converted to Indicated Mineral Resources.

Adjustments to the Mineral Resources are supported by the following:

• 57 additional drill holes provide an increase in data density; • Re-interpretation of mineral deposit structure and geology that has been applied to the Mineral Resource estimation process; • Adjustment to the cut off applied to Mineral Resource. Cut off grade used is 6% Zn, which is based on mining studies. The previous cut off was 10% combined Pb+Zn. • 5.7% reduction in density applied to the Mineral Resource based on recent measurement and analysis of drill core density.

AMC Consultants Pty Ltd provided technical expertise in the estimation of the Dugald River Mineral Resources and reviewed the quality assurance and quality control of the data.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 70 of 132 Mineral Resources 2008 Explanatory Notes

Figure 10 Location of Dugald River Project

6.2 Geology

The Dugald River zinc, lead and silver deposit is located 65km north west of Cloncurry, Queensland. It is situated within complexly deformed and metamorphosed mid Proterozoic sediments of the Eastern Succession in the Mt Isa Inlier.

The host Corrella Formation, an inter bedded sequence of greenschist to amphibolite metamorphosed carbonate and siliclastic lithologies, occurs in a 4-kilometre wide, north-south trending high strain zone that is unconformably boarded to the west by the younger Knapdale Quartzite and to the east by the Mt Rosebee Fault.

Dugald River mineralisation occurs as a westerly dipping (45-85º) sphalerite, galena, pyrite and pyrrhotite rich horizon which is hosted within a 300m thick package of black slates. It has a strata bound tabular geometry, three kilometres in strike, between 1 and 30m thick and extends over one kilometre in depth. Shearing along the mineralised horizon causing brecciation and re- mobilisation of sulphides has resulted in a mix of primary laminated and primary and secondary vein and breccia textures.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 71 of 132 Mineral Resources 2008 Explanatory Notes

Various genetic models have been proposed for the metal sulphide accumulation, including deposition from migrating basin brines during diagenesis to much later syn-tectonic deposition driven by metamorphic processes.

6.3 Assessment and Reporting Criteria

The following table provides a summary of important criteria related to the assessment and reporting of the Dugald River zinc, lead and silver Mineral Resource.

Criteria Status Sampling Techniques and Data

Drilling techniques • NQ, NQ2, HQ, BQ and PQ diamond and Reverse Circulation drilling techniques have been employed between 1939 and 2008. Sampling completed before 1969 was not used in this Mineral Resource estimate.

Drill sample • Drilling recovery for all practical purposes within the Mineral recovery Resource zone is approximately 100%. Recovery is recorded in the database.

Logging • Standard geological logging of lithology, alteration, mineralisation, geotechnical parameters, magnetic susceptibility and structure (Ezy- Mark core orientation system) were completed. • The majority of mineralised intervals have been photographed. • Logging has been recorded on paper and Excel spreadsheets. • Core is stored at the Dugald River and Century Mine core processing facilities.

Sub-sampling • 2007/8 sampling procedures involved optimal 1m lengths (0.7-1.3m) techniques and of core being halved using a circular diamond saw, density tested sample and despatched to ALS laboratory Townsville for elemental analysis. preparation • Various sub-sample techniques and sample preparation techniques were used prior to 2007 including whole, ¾, ½ and ¼, core. Sample length in general is 1 metre, some 2 metre RC samples are in the dataset.

Quality of assay • 2007/2008 laboratory sample preparations involved drying, crushing data and and pulverising the entire metre of ½ core to 85% passing 75μm. laboratory tests Analysis was completed by ICP-AES following a four acid digest on samples over 1% zinc or lead or over 25ppm silver. • 2007/8 quality control practices included the use of standard, blank and duplicate sampling. Generic and custom standards (including manganese certified) were used at a rate of 2 in 30, field duplicates and blanks at a rate of one per hole. • AMC Consultants Pty Ltd completed a review of the database QC concluding that the data is suitable for Mineral Resource estimation.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 72 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Verification of • In 1997 a re-assay program of the majority of pulps greater than 5% sampling and zinc was completed. In general, zinc results showed good assaying repeatability. Results from the 1997 re-assay program were used in this estimate. • Re-assaying of umpire pulps were undertaken in the 2007/8 program. • Field duplicates and various forms of hole twining have being used.

Location of Data • An independent commercial surveyor has surveyed all locatable points collars using differential GPS and conventional survey techniques between the years of 1994 and 2008. • Down-hole surveying using Eastman, Reflex and gyroscopic cameras has been employed. Generally surveys are at 30m spacing down hole. • Intermittent interference due to the effect of magnetite and pyrrhotite required some individual non-gyroscopic surveys to be removed from the dataset. • The majority of holes drilled in 2007/8 were gyroscopically surveyed.

Data spacing and • The Mineral Resource was calculated on 290 drill hole intercepts distribution including 57 drilled during the 2007/8 program. • The Measured Mineral Resource was defined over the central portion of the mineralised zone, up to 1000m in length by 700m depth and has a variable hole spacing of 25-75m. • On the margin of the measured zone the indicated Mineral Resource was defined over an area spaning 2000m in length by 1000m in depth. This area has a variable hole spacing of 50-200m. • On the margin of the indicated zone the Inferred Mineral Resource was defined over an area spaning 2500m in length by 1000m in depth. This area contains has a hole spacing of 100-300m.

Orientation of • The mineralised zone strikes north south and dips at 85º-45º to the data in relation to west. Drilling is generally from west to east, variation of drilling geological angle has intercepted the mineralised zone and strata between 90º structure and 45º to strike and between 10º and 90º to inclination. Estimating and Reporting of Mineral Resources

Database integrity • IO Global has managed the Dugald River database between 1999 to the present. A reconfiguration of the database took place in late 2007 to early 2008. • The database is a semi-normalised, fully relational data system specifically designed for exploration and mining drill and surface geochemistry and metadata. The database utilises the Microsoft SQL 2000 platform. Data is constrained in the tables by primary and foreign keys as well as a series of stored procedures and triggers.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 73 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Geological • The mineralised zone is modelled within two domains. Geological interpretation logging captures texturally different combinations of sphalerite, galena, and gangue pyrrhotite, pyrite, and slate. The inner domain broadly defines a continuous horizon of massive and breccia sulphide textures. The outer domain defines the surrounding lower grade vein sulphide and shoots of discontinuous massive and breccia sulphide mineralisation. • The dominant structural fabric runs parallel to the mineralised zone. In some regions faulting cuts the mineralised horizon. A shallow west dipping structure set influences the mineralisation geometry.

Dimensions • The defined Mineral Resource is a sub-vertical tabular zone striking north-south on mine grid and inclined to the west. • The deposit outcrops at surface and has a strike length of 2400m between 13350N and 15750N. In the north the mineralised zone dips at 85º to the west. Near surface at 14400N the mineralised zone shallows in dip for a strike of 200m, which forms a flexure that plunges at 40º to the south-west. The dip of this zone shallows to 45º at 500m below surface. Further south the flexure opens and the dip increases. • The true thickness of mineralisation is up to 30m. The thickest zone occurs in a north plunging shoot 200m in strike near the surface at 14000N that broadens to 600m strike at 800m depth.

Estimation and • In general a 2% zinc envelope with edge grade carrying any dilution Modelling to greater than 2% was used for the outer domain and a 9% zinc Techniques envelope with edge grade carrying any dilution to greater than 9% was used for the inner domain. • Ordinary Kriging was used to estimate zinc, lead, silver and manganese separately in the inner and outer mineralized domains and iron within the outer domain only. An octant search requiring a minimum of two octants with sample data for a block estimate was used. Drill samples were composited to 1m. Silver was top cut at 600ppm.

Moisture • The density of core was measured at atmospheric conditions generally within two weeks of drilling. Tonnage based on this density does not contain in situ moisture content. • Moisture content is considered minimal.

Cut-off • For the optimal definition of zinc metal, which comprises the parameters majority of the deposit value, a zinc only cut-off has been employed. • The cut-off of 6% zinc is used as it represents the mineralisation with realistic prospects of extraction as defined during investigations for the 2008 feasibility study.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 74 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Mining factors or • Assumptions include that conventional underground long hole assumptions stoping will be employed. No minimum Mineral Resource thickness was used. Zones of internal dilution are not reported as part of the Mineral Resource.

Metallurgical • Assumptions include that conventional flotation and recovery of zinc factors or and lead will take place, with silver reporting to the lead concentrate. assumptions The zinc concentrate will include a level of manganese impurity that is marketable and treatable.

Bulk density • 1,898 1 metre ½ NQ core samples from the 2007/8 drilling program were tested on site using the wet and dry weight method of density determination following the Archimedes Principle. No sealants were employed, as core has low porosity and does not readily absorb water. • A stoichiometric bulk density formula was calibrated using the results. These samples represent approximately 25% of all drilling through the mineralised zone. • The stoichiometric formula produces a resultant bulk density for model blocks by combining the variables of iron, zinc and lead from estimated block grades with the calibrated constants representing the density of sphalerite, galena, iron sulphide and gangue.

Classification • The Measured, Indicated and Inferred Mineral Resource is reported at a 6% zinc cut off. This best defines the mineralisation that the JORC statement “reasonable prospects for eventual economic extraction” refers to when the “technical and economic factors” defined in the 2008 feasibility study are considered. • The Mineral Resource is classified as Measured, Indicated and Inferred based primarily on drill hole spacing but also on geology and data quality considerations. • Conservatively less than 50m was extrapolated beyond the drilling margins.

Audits or Reviews • Behre Dolbear Australia (BDA) completed a review of the Mineral Resource estimate.

• AMC Consultants completed a review of the sampling database and quality control and assessment to August 2008. They recommend that the data used is suitable for Mineral Resource estimation. AMC Consultants also provided technical expertise in the Mineral Resource estimation process.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 75 of 132 Mineral Resources 2008 Explanatory Notes

6.4 Competent Person Statement

The information in this report that relates to Mineral Resources is based on information compiled by Mr Peter Carolan, who is a Members of the Australasian Institute of Mining and Metallurgy and has sufficient experience which is relevant to the style of mineralisation and type of deposit under consideration and to the activity which they are undertaking to qualify as Competent Persons as defined in the 2004 Edition of the ‘Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves’. Mr Peter Carolan consents to the inclusion in the report of the matters based on their information in the form and context in which it appears.

Mr Peter Carolan is a full-time employee of OZ Minerals Limited.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 76 of 132 Mineral Resources 2008 Explanatory Notes

7 GOLDEN GROVE MINERAL RESOURCES STATEMENT - 30 JUNE 2008

7.1 Summary

The Golden Grove Mineral Resource compiled by OZ Minerals in accordance with the ‘Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves’ (JORC Code) 2004 edition, was updated in June 2008.

The Mineral Resource Statement incorporates primary zinc, primary copper, oxide copper and oxide gold Mineral Resources within the Scuddles and Gossan Hill mines at the Golden Grove Operations. The Mineral Resource includes Inferred, Indicated and Measured categories (Table 16) and is inclusive of the quoted Reserve.

Table 16 Golden Grove Mineral Resource as at 30 June 2008.

Mine Resource Commodity Tonnes Cu Pb Zn Ag Au Category (Mt) % % % g/t g/t Gossan Hill Measured Primary Zinc 3.43 0.4 1.4 13.5 89 1.9 Primary Copper 5.88 3.1 0.1 0.6 19 0.7 Oxide Copper ------Oxide Gold ------Indicated Primary Zinc 1.31 0.5 1.9 13.6 110 2.2 Primary Copper 3.67 2.9 0.0 0.4 16 0.6 Oxide Copper 4.08 1.9 - - - - Oxide Gold 1.04 - - - 94 3.1 Inferred Primary Zinc 2.76 0.8 0.9 12.2 69 1.1 Primary Copper 5.79 3.3 0.0 0.9 28 0.8 Oxide Copper ------Oxide Gold 0.07 - - - 197 4.3 TOTAL 28.03 2.1 0.4 3.9 39 1.0 Scuddles Measured Primary Zinc 1.06 0.5 0.9 11 82 1 Primary Copper 2.89 3 0.1 0.9 19 0.5 Indicated Primary Zinc 0.12 0.5 0.9 10.1 71 0.7 Primary Copper 1.75 2.7 0.0 0.4 10 0.3 Inferred Primary Zinc 1.29 0.7 0.8 11.6 69 0.9 Primary Copper 1.27 2.6 0.0 0.2 14 0.2 TOTAL 8.38 2.17 0.28 3.75 32.77 0.54 Total Golden Grove Mineral Resource 36.43 2.1 0.3 3.8 37 0.9

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 77 of 132 Mineral Resources 2008 Explanatory Notes

NB: All Mineral Resources quoted are at a nominal economic cut-off of >= AUD70/t NSR for Primary Zinc and Copper. Oxide Copper report above a cut-off grade of 0.5%Cu and Oxide Gold has a 1.0g/t Au cut-off grade. Significant figures do not imply precision. Figures are round according to JORC Code guidelines.

The primary zinc and copper Mineral Resources have increased by 5.57 Mt, from 24.57 Mt to 30.14 Mt.

Table 17 Comparison of 30 June 2008 Total Mineral Resources with June 2007 Total Mineral Resources Estimate

30-Jun-08 30-Jun-07 Commodity Tonnes Grade Tonnes Grade (Mt) % (Mt) % Primary Zinc 9.99 12.6% Zn 9.36 12.8% Zn Primary Copper 21.25 3.0% Cu 15.2 3% Cu TOTAL PRIMARY 31.24 24.57 Oxide Copper 4.08 1.9% Cu 4.08 1.9% Cu Oxide Gold 1.11 3.2 g/t Au 1.11 3.2 g/t Au TOTAL OXIDE 5.19 5.19

NB: No new drilling results were available prior to the June 2008 Mineral Resource Estimation in the oxide copper and gold Mineral Resources areas therefore no changes are recorded between 2007 and 2008.

The increase in the primary zinc and copper Mineral Resource is due to a combination of factors:

• Addition of the Xantho Extended Inferred Mineral Resource at the Gossan Hill Mine. • Addition of the Cervantes Inferred Mineral Resource at the Scuddles Mine. • Re-evaluation of material previously deemed unrecoverable at the Scuddles Mine. • Adjustments to economic assumptions – this has a small impact in comparison to the above. • Mining depletion. • Re-interpretation of mineralisation models.

During the period from June 2007 and June 2008 significant changes were made to the geological models and block model estimation parameters. The impact these changes had on the Mineral Resource estimations are discussed in detail in the Results section of this report.

7.2 Introduction and Setting

The Golden Grove June 2008 Mineral Resource Statement incorporates the primary zinc, primary copper, oxide copper and oxide gold Mineral Resources at the Scuddles and Gossan Hill deposits.

The Golden Grove operation is located 375 km NNE of Perth and 225 km due east of the coastal port town of Geraldton. Access to site is via sealed roads from Perth to Paynes Find and from Geraldton to Yalgoo.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 78 of 132 Mineral Resources 2008 Explanatory Notes

The Golden Grove operation is wholly owned by OZ Minerals Golden Grove Pty Ltd which is a subsidiary company of OZ Minerals Pty Ltd. Golden Grove was previously owned by owned by Oxiana Limited. Oxiana Limited merged with Zinifex Limited in 2008 to form OZ Minerals Pty Ltd.

The Golden Grove operation comprises of underground operations at Gossan Hill and Scuddles. VHMS mineralisation was discovered at Gossan Hill in 1971 and at Scuddles in 1979. Scuddles has been producing zinc and copper for over 14 years and Gossan Hill started producing in 1998. The Scuddles mine was placed on care and maintenance in July 2005 and was re-commissioned in 2007 by Oxiana Limited.

Figure 11 Golden Grove Location Plan

The two deposits are within the same stratigraphic horizon but 4km apart. The stratabound mineralisation is hosted in Golden Grove Member 6 (GG6) of the Golden Grove Formation. Gossan Hill does however also contain copper mineralisation associated with magnetite in GG4 stratigraphic unit. The mineralisation is massive to stringer style, with minor breccia mineralisation, continuous along strike and down dip.

The sequence is intruded by dolerite dykes, sub-vertical rhyolite dykes and a voluminous dacite body with associated feeder dykes which stope out portions of mineralisation.

At Scuddles the zinc and copper mineralisation is hosted conformably within the upper portion of GG6, with documented dimensions of >1000m down dip, 500m along strike and up to 50m in width. Late stage low-grade regional metamorphism, folding and faulting have resulted in the Scuddles mineral deposit dipping steeply at 70 to 80 degrees to the west.

The mineralisation at Scuddles occurs in a number of distinct lenses:

• Main Lens, Central Lens and Deeps Lens, which have contained the bulk of historically exploited mineralisation. Occur in the upper part of the deposit above 10630mRL.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 79 of 132 Mineral Resources 2008 Explanatory Notes

• The deep Cervantes zinc and copper mineralisation which extends below 10630RL. • Zeewijk Zinc, located below the Scuddles Fault at 9600mRL (850m depth) is up to 20m thick, 136m along strike and 140m high. • Batavia Zinc and Batavia Copper, which lie beneath a second fault known as the Batavia Fault. • The mineralisation at Zeewijk and Batavia is the shallower equivalent of the Cervantes mineralisation.

The new Cervantes copper and zinc mineralised zones occupy a similar stratigraphic position to the existing Scuddles zones above. The copper mineralisation is considered to be stratabound within the GG6 bedding orientation while the zinc appears to have a strike of about a 2o difference. The copper mineralisation trends at about 010o (local mine grid) in the southern part of the Mineral Resource area and swings around to approximately 020o in the north of the Mineral Resource. Separate stringer zones are parallel with this northern orientation. The main Cervantes copper body extends from 22375mN to 22530mN and from 8640mRL to 8960mRL. The maximum thickness of 23m occurs at around 8850mRL, reducing gradually towards the boundaries of the Mineral Resource model. The northern copper stringers extend from 22460mN to 22540mN and from 8800mRL to 8560mR L. They have general thicknesses of around 4m to 11.8m. Each has a bulge of approximately 11m at 8650mRL and 8730mRL. The high grade zinc mineralisation trends at approximately 008o, interpreted as cross cutting the copper to the north. This mineralisation remains at an approximately constant thickness of 3.2m . The zinc mineralisation in the current Mineral Resource model extends from 22350mN to 22530mN and from 8600mRL to 8960mRL with the upper northern extent retreating southwards with reducing depth to 22450mN at 8960mRL.

At Gossan Hill, weathering and oxidation extend down approximately eighty metres from surface. Oxide/supergene copper and oxide gold Mineral Resources are located directly above the primary copper and zinc mineralisation respectively.

The oxide copper mineralisation is the weathered portion of the primary copper mineralisation. It is hosted within weathered GG4 sediments, a flat lying dolerite and sub vertical gossans. The mineralisation occurs from 18240mN to 18640mN, is up to 70m wide and extends to the base of weathering at approximately 80m to 100m depth. The mineralisation, which comprises predominantly malachite and azurite, continues through the base of weathering into the primary zone.

The gold mineralisation continues below the base of weathering into the primary zinc mineralisation. Gold occurs as discrete grains (1-4 microns) locked within the ironstones and is associated with elevated silver. The oxide gold mineralisation extends from 17720mN to approximately 18300mN, is up to 20m wide and extends from surface to the base of weathering at approximately 80m below surface.

The main primary copper zone, known as the A Copper and Q Copper zones, extends 700m along strike, 450m down dip and 80m wide. Primary mineralisation occurs as chalcopyrite with various gangues. The system is broadly differentiated into a footwall massive sulphide zone, grading into a magnetite - sulphide zone, with stringer-style mineralisation throughout. Copper grades

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 80 of 132 Mineral Resources 2008 Explanatory Notes transgress sulphide/magnetite lens boundaries. A massive magnetite zone occurs in the hanging wall to the mineralisation.

Significant copper also occurs in the footwall of the GG6 zinc mineralisation in C Zinc, Hougoumont and to a lesser extent at Amity. This mineralisation comprises stringer to sub-massive chalcopyrite and pyrite with or without magnetite.

The main zinc mineralisation at Gossan Hill is stratabound within the GG6 unit. The main sulphide types are pyrite, sphalerite, chalcopyrite, pyrrhotite and galena, with minor tetrahedrite, arsenopyrite, cassiterite and bismuthinite.

Mineralisation is divided into several zones.

• The A and B Zinc zone is 280m along strike, 300m down dip, 20m to 30m wide and tabular. A and B Zinc are two separate centres of mineralisationn that did not coalesce, even though they occupy the same stratigraphic level. • C Zinc is interpreted to occupy a fold hinge in a more thickly developed portion of GG6, or to represent the primary sulphide mound that has been folded. It forms a pipe-like structure plunging steeply to the south, measuring approximately 50m across and 250m down dip. Copper and zinc occur in a footwall chalcopyrite and sphalerite stringer zone. • Amity mineralisation lies approximately 400m below C Zinc and beneath the Main Aquifer Fault. The main Amity zone is a pipe like body within the same fold structure that hosts C Zinc. Mineralisation is open at depth. • Hougoumont lies to the north of Amity, separated by a steeply dipping dacite intrusive and terminated to the north by a further dacite intrusive. Two main sphalerite lenses occur, in GG6 and in the SC2 horizon in the hanging wall. The mineralisation in SC2 is high in zinc, lead, gold and silver and very low in iron. • Catalpa and Ethel lie approximately 500m below surface and 1,000m North of Amity and are terminated at the northern end by the Catalpa Fault. • Xantho occurs north of the Catalpa Fault and 380m below Catalpa. • Xantho Extension is believed to be the down dip equivalent of Xantho. The mineralisation is constrained to the south by the Catalpa Fault which strikes NE/SW and dips 85o towards the east. The fault position and orientation below the 8900mRL is poorly understood and requires further investigation through drilling and re-logging of holes. Alternate interpretations of the fault position could potentially have implications on the lateral and vertical continuity of the model. The northern boundary of the mineralisation coincides with an interpreted offset in the stratigraphy. This could indicate a northern structural control on the mineralisation and potentially have implications on the continuity of the model • Cambewarra is a northern extension of Catalpa displaced by the Catalpa Fault and moved 200m to the east.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 81 of 132 Mineral Resources 2008 Explanatory Notes

7.3 Results

Mineral Resources have been estimated using the Golden Grove geological database as at 1 June 2008. The majority of the Mineral Resources comprise primary zinc and copper mineralisation in and around the Scuddles and Gossan Hill underground operations. The near surface oxide gold and copper Mineral Resources located at Gossan Hill were estimated in 2003 (oxide gold by Newmont Australia Limited) and 1997 (oxide copper by Normandy Mining Limited), and were not updated for this estimate.

Mineral Resource block modelling for primary zinc and copper mineralisation has been reviewed and changed from previous years; these changes are outlined in this Statement and are discussed in detail in the full Ore Reserve and Mineral Resource estimation report.

Geological domains, generated using drill information and mine development where available, served as constraints for block modelling and grade estimation. Estimation of grades and densities into block models were carried out using ordinary kriging for the majority of Mineral Resources and inverse distance squared where information was sparse.

The Mineral Resource includes Inferred, Indicated and Measured categories and is inclusive of the quoted Reserve. The Mineral Resource estimate has been classified based on data density, data quality, confidence in the geological interpretation and confidence in the estimation.

The primary zinc and copper Mineral Resources were prepared by using the mined out void shapes within the initial block model generation. The mind out void shapes were translated east and west, also up and down in the Scuddles block model, to flag non-recoverable material adjacent to stoping. All mined and non-recoverable flagged blocks were excluded from the Mineral Resource estimate. Within Scuddles further areas considered unmineable due to position (along hanging wall of previously stoped areas) or too small in tonnage to be potentially economic were also excluded from the Mineral Resource estimate. These unmineable areas were determined in consultation with the Senior Mine Engineer.

The cut-off grade used for the primary zinc and copper Mineral Resource estimate approximates the site mining and processing break-even costs, taking into account metallurgical recovery, concentrate transport costs, concentrate treatment and refinery charges and royalties. Expressed as Net Smelter Return (NSR) or mine gate value, the cut-off grade used for the Mineral Resource estimate is AUD70.00/t.

The metal prices and exchange rate used for the estimation of the Mineral Resources are presented in Table 18 below.

Table 18 Metal Prices and Exchange Rate

Metal Unit Price Zinc USD/lb 1.00 Copper USD/lb 2.20 Lead USD/lb 0.50

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 82 of 132 Mineral Resources 2008 Explanatory Notes

Gold USD/oz 700 Silver USD/oz 11.00 Exchange rate AUD/USD 0.75

The cut-off used for the oxide gold Mineral Resource estimate was 1g/tAu. The cut-off used for the oxide copper Mineral Resource estimate was 0.5%Cu.

Table 19 compares the June 2008 Mineral Resources with the previously reported Mineral Resource at 30 June 2007.

Table 19 Comparison of 30 June 2008 Mineral Resources with June 2007 Mineral Resources Estimate (Measured + Indicated + Inferred)

30-Jun-08 30-Jun-07 Tonnes Grade Tonnes Grade (Mt) % (Mt) % Primary Zinc 9.88 12.5% Zn 9.36 12.8% Zn Primary Copper 20.84 3.0% Cu 15.2 3.0% Cu TOTAL PRIMARY 30.72 24.57 Oxide Copper 4.08 1.9% Cu 4.08 1.9% Cu Oxide Gold 1.11 3.2g/t Au 1.11 3.2g/t Au TOTAL OXIDE 5.19 5.19

The primary zinc and copper Mineral Resources have increased by 6.15 Mt, from 24.57 Mt to 30.72 Mt. The increase in the primary zinc and copper Mineral Resource is due to a combination of factors:

• The addition of the Xantho Extended Inferred Mineral Resource at the Gossan Hill Mine, 4.7 Mt. • The addition of the Cervantes Inferred Mineral Resource at the Scuddles Mine, 1.64 Mt. • Re-evaluation of material previously deemed unrecoverable at the Scuddles Mine, an additional 1.0Mt. • Mining depletion, -1.45 Mt. • Re-interpretation of mineralisation models. • Sterilisation of areas due to drilling. • Changes to block model grade estimation parameters. • Adjustments to economic assumptions – The script used to populate the dollar variables in the block model was reviewed, simplified and changed. This new script more accurately reflected all the costs associated with mining, milling, freight and smelting. Comparison

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 83 of 132 Mineral Resources 2008 Explanatory Notes

between this year’s economic assumptions and last year’s show that increasing metal prices had very little effect on total Mineral Resources at Scuddles, an increase of 0.08 Mt, but a noticeable change in the Gossan Hill Mineral Resource, with an increase of 1.41 Mt.

7.4 Geological Interpretation, Estimation and Modelling Techniques

All Mineral Resource modelling was carried out using VulcanTM software. Solid triangulations were produced around mineralised envelopes within which grade was interpolated.

• Primary zinc mineralisation triangulations were based on a >4% zinc cut-off. • Primary copper mineralisation triangulations were based on >0.2% copper cut-off. • High-grade copper mineralisation triangulations were generated for certain areas of the mine to reduce the smearing effect of unconstrained high grades and were based on a 1%Cu or in places 2%Cu cut-off. • Grades below these cut-offs were included in some areas to honour the geology.

The oxide gold deposit is located above the primary zinc mineralisation within the oxidised massive sulphide lithology in the area where zinc was leached out. The boundary of the oxidised massive sulphide corresponds with a 0.2g/t Au cut-off and this cut-off was used to constrain the interpreted gold mineralisation.

Numerous un-mineralised dacite, dolerite and rhyolite intrusive rocks that cut across the mineralisation were also modelled.

In the block modelling process, each domain was coded with a unique number or rcode, with the priority set to the same as the rcode value to overcome overlapping triangulations. The rcode was also used to code drill hole samples from that domain. All of these rcodes were tabulated with the name and description of the corresponding triangulation.

Samples were composited to 1m lengths in the primary zinc and copper deposits and to 2m in the oxide deposits, which corresponds to the average sample length. Top-cuts were only applied to gold in one of the Amity massive pyrite areas and one of the B Zinc massive sulphide areas.

The predominant method of estimating grade and bulk density into the block model areas, defined by the mineralisation triangulations, was ordinary kriging. All variography for the Gossan Hill Mine was reviewed and updated at the end of 2007 to include all drilling results obtained since the last variography review in 2005. Several areas of the Gossan Hill mine that were being estimated with inverse distance squared estimation are now being estimated with ordinary kriging as a result of this review. Where insufficient data did not permit geostatistical methods to be used, the inverse distance squared method of grade estimation was used. The Gossan Hill oxide copper and gold Mineral Resources were estimated using indicator kriging.

During the period between June 2007 and June 2008, significant changes were made to the geological interpretation triangulations and block model estimation parameters for the primary zinc and copper mineralisation which have affected the Mineral Resource estimate and

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 84 of 132 Mineral Resources 2008 Explanatory Notes consequently the ore reserve. All of these changes are discussed in depth in the Mineral Resource and Ore Reserve estimation report but are briefly discussed here.

Geological interpretation, estimation and modelling techniques vary between the Scuddles and Gossan Hill mines and are therefore discussed separately.

7.5 Changes to Estimates

7.5.1 Changes made to both Gossan Hill and Scuddles

• All validated drill hole information was honoured. Any suspect holes were tabulated and removed from the Mineral Resource estimate. • Parent block size for all models was set to 20m x 50m x 50m. • Block size limits within mineralisation interpretations were changed to 5m x 5m x 5m to match half the drill hole spacing. • Sub-cells were set at 1m x 2.5m x 2.5m for definition of contacts. • Parent block estimation ensured that sub-cells were assigned the same value as the parent cell. • All block models were created with all variables included in the initial block modelling process and not added prior to the initial block model creation. • Mined void shapes were updated to include all mined out stopes and development. • Mined voids were blocked in the initial block modelling process. • Triangulation priority codes (rcode variable) were changed to allow the intrusive bodies to sterilise mineralisation where these triangulations intersect each other. • The number of estimation runs and the requirements used to calculate each run for all block models was standardised. - 4 estimation runs were used for all variables - Discretisation was set to 4m x 4m x 4m. • The economic assumptions script was reviewed, updated and simplified. • All stope shapes created in the reserve estimation process were geologically assessed and given a geological risk rating. Where necessary, Mineral Resource categories for the individual stopes were changed and updated in the block models. This information has been tabulated and is located in the public access drive for all Mine Engineers and Geologists to view. These stope evaluations also provide the basis for the Stope Design Records written for all stopes in the mine schedule.

7.5.2 Changes to Gossan Hill

• The ABCD Zinc Block model did not change from the previous Mineral Resource estimate, new economic assumptions were run on the old model and figures quoted. New economic assumptions were run on the old model and figures quoted.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 85 of 132 Mineral Resources 2008 Explanatory Notes

• The block model encompassing the Camberwarra and Xantho mineralisation, the Gossnorth block model, was not changed from previous estimates. New economic assumptions were run on the old model and figures quoted. • Geological triangulation interpretations were formed from polygons snapped to drill holes. Geological modelling included mineralised lenses, using a 4% Zn cut-off for zinc mineralisation and a 0.2% Cu cut-off for copper mineralisation, intrusive bodies and hanging wall barren units terminating the mineralisation. • Introduction of high grade 1% copper triangulations where appropriate to avoid smearing of high grade copper into the existing 0.2% copper triangulations. • All variography was reviewed and updated to include drilling data obtained since the last review in 2005. • Upgrade of areas to kriging estimates where information allowed – Hougoumont, Catalpa and Ethel areas. • The number of estimation runs and the requirements used to calculate each run for all block models was standardised. - 4 estimation runs were used for all variables. - Octant based searches with addition constraints were included in the grade estimation calculations. - Search ellipse distances and directions were modified. - Maximum and minimum samples used per estimate were modified. - Discretisation was set to 4m x 4m x 4m.

Table 20 Block Model Estimation Parameters

Parameter Run 1 Run 2 Run 3 Run 4 Sample and Octant Parameters Samples per estimate Min 10 10 5 2 Max 15 15 15 15 Octants with samples 5 5 4 - Samples per Octant Min 5 5 5 - Max 1 1 1 -

• Search ellipse dimensions were roughly based on drill spacing, doubled for runs 2 and 3 and tripled for run 4. • For low grade zinc and copper domain shapes, search parameters were relaxed. • In areas of the mine in which the triangulations were not updated, historic search ellipse dimensions were retained. Triangulations were not updated in all areas for several reasons: - No new drilling in areas; - No stoping in areas scheduled in the near future;

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 86 of 132 Mineral Resources 2008 Explanatory Notes

- No active mining in areas.

• Most of the triangulations defining Mineral Resource category were updated as required. • The post block model scripting to populate “Orebody” and Area was updated to include all rcodes and was simplified.

7.5.3 Changes to Scuddles

• Geological triangulation interpretations were formed from polygons snapped to drill holes for areas drilled post June 2007. Geological modelling included mineralised lenses using a 4% Zn cut-off for zinc mineralisation and a 0.2% Cu cut-off for copper mineralisation, intrusive bodies and hanging wall barren units terminating the mineralisation. • Suspect holes were identified, tabulated and removed from the Mineral Resource estimate. • All block models were created with all variables included in the initial block modelling process and not added prior to the initial block model creation. • Mined void shapes were updated to include all mined out stopes and development. • Mined voids were blocked in the initial block modelling process. • Triangulation priorities (rcode variable) were changed to allow the intrusive bodies to sterilise mineralisation where triangulations intersect each other. • Sub-block size limits within mineralisation were changed to match half drill spacing widths. • The number of estimation runs and the requirements used to calculate each run for all block models was standardised: - Octant based searches with addition constraints were included in the grade estimation calculations. - Search ellipse distances and directions were modified. - Maximum and minimum samples used per estimate were modified. • The post block model scripting to populate Orebody (previously region, but changed to fall in line with Gossan Hill terminology) and Area was updated to include all rcodes and was simplified. • The economic assumptions script was reviewed, updated and simplified. • All stope shapes created in the reserve estimation process were geologically assessed and given a geological risk rating. Where necessary, Mineral Resource categories for the individual stopes were changed and updated in the block models. This information was tabulated and is located in the public access drive for all Mine Engineers and Geologists to view. These stope evaluations also now provide the basis for the Stope Design Records written for all stopes in the mine schedule.

7.6 Classification

Mineral Resources were classified based on data density and quality. The drill density required for each area was assessed based on the geology and continuity of mineralisation in that area. The estimation run in which the block was filled together with the number of drill holes and number of

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 87 of 132 Mineral Resources 2008 Explanatory Notes

samples used in the estimate were also taken into account when determining the classification. The requirements used for each classification is presented in the table below.

Table 21 Mineral Resource Classification Criteria

Number Estimation Number of Classification Hole Spacing of Drill Run Filled Samples holes Measured 10mx10m to 15mx15m 1-2 >=5 10-15 Indicated 20mx30m to 30mx30m 2-3 2-5 5-15 Inferred Wider spacing 3-4 <=2 1-15

Knowledge of the deposit including the complex nature of the numerous intrusive bodies was also taken into account when determining the classifications.

Further to this, all stopes were individually assessed to determine if the correct confidence had been allocated and the geological risk was also rated for each stope. A table showing a break down of the stope shapes by drill spacing, assay and block accuracy, confidence, geology, structure, associated mapping, proposed drilling and any other geological concerns was completed.

7.7 Assessment and Reporting Criteria

The following tables provide a summary of important criteria related to the assessment and reporting of the Golden Grove Mineral Resource.

All of the information in the following table refers to criteria applied to both the Scuddles and Gossan Hill Mineral Resource estimations.

Criteria Status Sampling Techniques and Data The Golden Grove drilling database comprises 1,769,048m of surface (near mine and exploration) and underground drill meters. Several methods of drilling have been carried out on the Golden Grove leases over the years, consisting of 71% diamond core, 24% RAB drill holes, 4% RC holes and minor aircore holes. Only diamond drill core and minor RC data was used in the Mineral Resource Drilling techniques estimations for Gossan Hill. Long surface, RC and steep angled holes that had become redundant as a result of new underground grade and delineation drilling were removed from the Mineral Resource estimation. These holes were used as a guide only in the modelling process. The total number of diamond drill holes used in the Mineral Resource

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 88 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status modelling is 2,627 for Scuddles and 3,930 for Gossan Hill. Since the previous Mineral Resource estimate at 30 June 2007, 392 drill holes were drilled totalling 32,614m of drilling within the Scuddles and Gossan Hill Mines. 30,380 m of underground diamond drilling was completed at Gossan Hill: 21,036m (288 holes) were designed to upgrade Indicated Mineral Resources to Measured Mineral Resources. 9,344m (83) were designed to upgrade Inferred Mineral Resources to Indicated Mineral Resources. 623m were drilled to obtain geotechnical data in unmined and un-drilled areas of the mine. 1,610m (18 holes) were drilled underground at Scuddles to upgrade Indicated Mineral Resources to Measured Mineral Resources within the zinc mineralisation areas around the 600 Level and the 845 Level. 23,426m were drilled into exploration targets to try and upgrade exploration results into an Inferred Mineral Resource. 10,517m (17 holes) were drilled into the Xantho Extension target at Gossan Hill. 10,612m (18 holes) were drilled into the Cervantes target at Scuddles and 2,297m (3 holes) were drilled into the Hougoumont extension exploration target, this area is still being drilled and is not included as part of the Mineral Resource estimation. There are also 348 Gossan Hill RC holes in the database, of which 267 holes were used in the modelling of the gold oxide and copper oxide deposits. Only 81 RC holes were used for modelling of Gossan Hill primary deposits and none to model the Scuddles deposit. Core recovery is measured between drill runs and recorded in the database. Core recovery is generally excellent for all core sizes, averaging 98% recovery. Drill sample Core loss blocks are placed by drillers as required at the time of drilling. recovery No bias for core loss in mineralisation is observed. Recoveries from RC drilling were reported to be generally good except for lower recoveries reported for drilling in parts of the gossans. All drill core is geologically logged using codes set up for direct computer input into an EXPLORER3 database. This database is currently being replaced by GBIS. Historic holes are being validated with the original data before being moved across to the new system. Holes drilled after June 2008 are Logging entered directly into GBIS. Rock type, stratigraphy, grainsize, weathering, colour, bedding, alteration style, type and intensity, structure, mineralisation style and percentage are recorded. Zones of sulphide mineralisation are assayed with at least 8m of waste on

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 89 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status either side of these zones. All RC chips were geologically logged. Prior to the 1998/99 drill programme and the subsequent Mineral Resource estimation, a programme of drill hole re-logging was carried out on all of the holes previously drilled into the oxide gold deposit. All diamond drill holes since 2003 have been photographed. Diamond drill holes are stored within the core yard facility. Assay pulps are stored for delineation and exploration programs in the core yard facility.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 90 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status All NQ diamond drill holes are ½ core sampled. LTK60 core was ½ core sampled up until January 2008 when the decision was made to whole core sample to get a better representative sample. Core is orientated along the apical trace of the reference plane ellipse (typically bedding) and then half cored using a diamond core saw to ensure the sample is a true representative of the in situ mineralisation. The optimal sample interval is 1m. Sample sizes can range from 0.5m to 1.5m. Sample intervals do not cross lithological boundaries. Sample density was measured using the gravimetric method. Samples at the off site commercial lab are dried, crushed and pulverised to a nominal 90% passing 75 microns. RC drilling samples were used extensively in the oxide deposits. Various sampling techniques were used for the RC drilling, which were most commonly sampled on 1m intervals in mineralised zones and as 4m composites outside of mineralised zones. Sub-sampling Prior to 2001 a number of commercial analytical laboratories were used for techniques and drill hole sample analysis. sample preparation Since 2001, ALS Chemex at Malaga, WA has analysed the majority of samples. Exceptions to this were the 2005 and some of 2006 Hougoumont drill samples were analysed by Ultra Trace Analytical Laboratories in Canning Vale, WA. Samples undergo total pulverisation before an assay charge is extracted and analysed for a basic suite of seven elements (Zn, Cu, Pb, Fe, S, Ag and Au). At ALS Chemex, the methodology for Zn, Cu, Pb, Fe, S, and Ag assay involves digesting a 0.25g pulverised sample in an aqua-perchloric acid solution and then testing it using inductively coupled plasma (ICP) techniques. At Ultra Trace, Zn, Cu, Pb, Fe, S and Ag are analysed by sodium peroxide fusion, hydrochloric digest, ICPOES determination and Au by 40g Fire Assay, ICPMS determination. For gold, a nominal 50g sample of the pulverised sample is mixed with fluxing agents and fused at high temperatures. The gold is dissolved in aqua regia and the solution analysed by atomic absorption spectrometer. Sampling procedures at the Golden Grove mine are deemed appropriate for the style of mineralisation.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 91 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Before 1997, umpire assays were used to assess the accuracy of assays. In the oxide deposits assay quality control was performed using standards, blanks, duplicates and umpire assays. Since 1997 and before January 2008, one standard per hole was submitted to the Laboratory for analysis. These QAQC samples represented approximately 4% of the database. Since January 2008 a standard, of suitable grade, is inserted every 20th sample and as the last sample of the hole. Results for field standards indicate no significant bias with only slight low bias in lead and zinc values (up to a relative 6%). Several batches have been re-analysed as a result of poor standard results. No duplicate samples were taken. Prior to January 2008 no field blanks were sent from site to the laboratory. This is now routinely carried out as part of QAQC procedures applied at Golden Grove. One blank is inserted for every 25 samples. Laboratory blanks generally return very low values with few infrequent higher values indicting minor contamination. In the oxide gold deposit the umpire assays results for the samples from the 1995 and 1998/99 drilling programmes show good correlation with original Quality of assay assays. Coarse crush and pulp repeat Au analyses of selected 1998/99 data and laboratory samples showed good correlation with the original assay, suggesting a low tests nugget effect for this mineralisation. For silver the mean value of the total for

all batches agrees very well between the original assay and the repeat. Details of assay validation on earlier programmes are not available. In the oxide copper deposit the results of various quality control programs are variable and it is difficult to draw meaningful conclusions based on the available data. OZ Mineral’s broad conclusions are that oxide copper assays are more likely to understate the true value than overstate them. A data validation and delineation drilling program began on the oxide copper and gold deposit in 2007. This is ongoing and findings will be available in time for the next Mineral Resource estimate. OZ Minerals employees carried out an annual Laboratory Audit of the ALS Chemex commercial Laboratory in Perth in September 2007. Findings of the audit were an overall good, however some areas for improvement were suggested: ALS should ensure that they test their “certified” chemicals rather than take them at face value. The sorting area needs to improve its process especially in light of recent lost samples. Records for calibration and tests should be readily available in each of the labs.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 92 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Verification of No umpire laboratory was used in the verification of any assays received sampling and between June 2007 and April 2008. assaying No twinned holes were drilled. All of the diamond drill-hole collar locations and orientations are surveyed using an electronic theodolite and recorded in either EXPLORER3 or GBIS. Down-hole surveying is performed using a Maxibor and/or an electronic multi-shot tool for all underground holes. Surface diamond drill-hole coordinates were surveyed in AMG using a DGPS method with a Leica GPS 900 instrument. This method gives an accuracy of 20mm. Surpac was used to convert AMG coordinates to local mine grid. All drill-holes have magnetic down-hole surveys taken at 30m intervals using a multi-shot down-hole camera. An azimuth adjustment of +52.5o degrees was applied for the conversion to local mine grid. Some surface diamond holes were partially surveyed by Downhole Surveys Pty Ltd using a HSHA Rate Gyro. Comparison of gyro and magnetic data showed slight differences in dip and azimuth readings. In general, magnetic data tended to be biased towards the north however the difference between the two has been determined to be negligible. Obtaining complete survey data for partially gyro-surveyed holes, involved adjusting multi shot magnetic data and adding this onto the last down-hole Location of Data gyro survey point. The adjustment value was calculated from the difference points between gyro and magnetic values at down-hole intervals of 30m. The difference in values is then averaged to obtain the adjustment. Adjustments to survey data other than conversion to mine grid was exclusive to holes surveyed with the gyro tool. Many of the surface RC holes had “in rod” dip only Eastman Surveys and a large number of these holes were not originally down-hole surveyed. Where possible, RC holes were subsequently re-surveyed using an electronic multi-shot tool. Holes containing magnetite were gyro-surveyed. Surface diamond drill holes have magnetic down-hole surveys taken at 30m intervals using a multi-shot down-hole camera. An azimuth adjustment of +52.5o degrees was applied for the conversion to local mine grid. Some surface drill holes were partially surveyed by Downhole Surveys Pty Ltd using a HSHA Rate Gyro. Comparison of gyro and magnetic data showed slight differences in dip readings and larger differences in the azimuth readings. In general, magnetic data tended to be biased towards the north, however the difference between the two has been determined to be negligible. All drill hole locations and orientations are validated by plotting collar pick

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 93 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status ups against existing surface or underground development and by comparing actual traces against designed drill-hole paths. All suspect holes have been flagged in the database and are removed from the Mineral Resource estimate database. Drill separation for copper mineralisation within both the Scuddles and Gossan Hill mines is approximately 10m along strike and up to 15m vertical for the Measured Category, double this spacing for Indicated and wider spacing for Inferred. Data spacing and Drill separation for zinc mineralisation within both mines is approximately 10- distribution 15m along strike and 10-15m vertical for the Measured category, double this spacing for Indicated and wider spacing for Inferred. Historically drill spacing was 20m along strike and 20m down dip for zinc mineralisation, however it was concluded that this spacing was too wide to accurately model geology to a Measured level of accuracy. All attempts are made to ensure that the drilling does not give a bias Orientation of data interpretation of the mineralisation. The procedure currently in place for in relation to designing drill holes states that the drilling angle to mineralisation is not to geological exceeded 30 degrees either side of the collar point, and 30 degrees vertical structure. when drilling. However, due to production and development constraints, this can not be followed on rare occasions. No audits or reviews have been carried out between the 2008 Mineral Resource estimate and the 2007 Mineral Resource estimate. An Internal review validating all data in the database is ongoing with the merger of all data into the new database. It is envisaged that this validation Audits or reviews will be completed for the next Mineral Resource estimation. QAQC procedures on site are being reviewed and new procedures are currently being formulated and put in place to come in line with the company QAQC standards.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 94 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Estimating and Reporting of Mineral Resources Database integrity All data is stored in the Explorer3 database or in the new GBIS database. All data is input manually into Microsoft Excel spreadsheets set up with key field and data validation processes. Conditional formatting functions and macros reduce input error. Only data signed off as valid on a data upload request form are loaded to the data base. Manual checks performed by the supervising geologist before data is uploaded to the database include; survey depth is less than end of hole depth, S.G min/max is within reasonable limits, RQD values are 100 or less, recovery files are within reasonable limits, and assays are checked against final hardcopies. Down-hole surveys are checked against each other and the proposed dip and azimuths. The survey department or contract survey companies pick up the collar co-ordinates of all holes which are again checked against planned. Underground drilling requires re-collaring if the first survey shot is +/- 1 degree out from design. All tonnages throughout the entire Mineral Resource, Ore Reserve and Moisture reconciliation process are reported as dry tonnes. Future mining factors and assumptions have been based on current mining practices. Mining comprises of long-hole open stoping. Ore is hauled to surface at Gossan Hill and in some areas of the Scuddles mine, which is determined by economic assumptions. The majority of ore mined at Scuddles is hoisted to surface using the existing Mining factors or hoisting shaft. assumptions Stope dimensions are approximately 5-10m wide, 25m high and 20m along strike for each panel. Stopes on average are 3 panels high. Dilution and recovery parameters are based on continual reconciliations and vary between areas of the mine. Investigations into the possibility of long-hole open stope mining the Hougoumont hanging wall narrow zinc mineralisation are currently being carried out. Metallurgical Metallurgical factors and assumptions are based on the current processing factors or plant producing copper, zinc and high precious metal (HPM) concentrates. assumptions Metallurgical recoveries are based on past performance. Copper recovery is approximately 80% and zinc is approximately 90%.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 95 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status All bulk density data collected from air dried drill core samples since 1997 (1m lengths or as specified for sampling) have been measured using a gravimetric method. As drill core samples are of solid, non porous fresh rock, the measured density is assumed to be the dry bulk density. Non-sampled intervals of waste rock are assigned an average density of 2.82 based on the mean density of analysed waste samples. Bulk density In the primary zinc and copper deposits, bulk density is interpolated within the block model using the same methods as for the grade estimation. In the oxide gold deposit bulk density was determined on selected holes using a gravimetric method. In the oxide copper deposit about 50% of samples have bulk in situ density measurements determined using a down-hole density probe. Average bulk densities were assigned in the block models of oxide deposits according to the rock type and depth below surface Mineral Resource block models were validated using various techniques. The sections and plans through the block models were visually compared to the drill-hole data. Block and sample statistics were compared for all domains. De-clustered statistical comparisons were also carried out for all domains. Block Model Sensitivities to minimum and maximum samples used, and interpolation Validation parameters were assessed during the estimation process. Waterfall diagrams were created for all block models showing the effect of changes made to the block models on the overall tonnages reported for copper and zinc. The June 2008 Mineral Resource estimate has been reviewed by Dan Grieg of Behre Dolbear in September 2008. The review focused on the interpretation of mineralised wireframes along with all other inputs and parameters. Conclusions from the review were: The current Mineral Resource modelling process at Golden Grove meets industry standards. Significant changes are directed towards developing a standard approach to both modelling and record keeping. Audits or Reviews Mineral Resource classification is believed to be conservative but in-line with mining requirements. Minor recommendations were made, however none of these will materially affect the quality of the Mineral Resource estimate. The last audit carried out on the Mineral Resource was by AMC in February 2006. Several of the recommendations have been put implemented in the construction of the 2008 Resource estimate, particularly in relation to simplifying the grade estimation files and post block model scripts.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 96 of 132 Mineral Resources 2008 Explanatory Notes

7.8 Competent Person Statement

Deposit Competent Person Primary Zinc and Copper Chevaun Gellie, OZ Minerals Golden Grove Pty Ltd Oxide Gold Robert Singer, Former Chief Geologist at Golden Grove (2003) (Gossan Hill oxide gold Mineral Resource Estimate as of October 2003, which remains unchanged for 30 Jun 2008) Oxide Copper Paul Blackney, formerly Snowden Mining Industry Consultants. (Gossan Hill oxide copper Mineral Resource estimate as of December 1997, which remains unchanged for 30 Jun 2008)

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 97 of 132 Mineral Resources 2008 Explanatory Notes

8 ROSEBERY MINERAL RESOURCE STATEMENT - JULY 2008

8.1 SUMMARY

The Mineral Resource at the OZ Rosebery Mine as at 30 June 2008 is summarised in Error! Reference source not found..

Table 22 Rosebery Mine Mineral Resource as at 30 June 2008

Category Tonnes Pb Zn Cu Ag Au Fe (Mt) (%) (%) (%) (g/t) (g/t) (%) Rosebery Measured 2.82 4.3 15.4 0.54 156 2.4 10.5 Indicated 3.91 3.4 11.3 0.37 132 1.8 9.4 Inferred 8.48 3.3 10.3 0.30 114 1.4 7.2 Meas+Ind+Inf 15.21 3.5 11.6 0.36 126 1.7 8.4

8.2 Introduction

The OZ Rosebery Mineral Resource Statement covers the zinc-lead-copper-silver-gold Mineral Resources for the OZ Rosebery Mine located in Rosebery, Tasmania, Australia. The known Mineral Resources are contained within numerous lenses, denoted by letters of the alphabet. The project is wholly owned by OZ Minerals.

This Statement is based on the annual Mineral Resources/Ore Reserves report produced at Rosebery (March 2008), with adjustments made to account for mining depletion between 31 March 2008 and 30 June 2008.

The Mineral Resource estimate was based on a geological database which included all the known Rosebery Mine drilling up to the present day. The geological interpretation, geological modelling, block model construction and grade estimation were completed on site by Rosebery Mine Geologists.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 98 of 132 Mineral Resources 2008 Explanatory Notes

8.3 Setting

Figure 12 Rosebery Location and regional geology

The OZ Rosebery Mine is a polymetallic (Pb-Zn-Cu-Ag-Au) deposit located at Rosebery on the west coast of Tasmania. The mineralisation comprises a series of massive sulphide lenses hosted within feldspar phyric, rhyolitic volcaniclastic sandstones and siltstones of the Cambrian Mt Read volcanic sequence referred to as the “Host Sequence”.

The Mt Read Volcanics are a 200 km long and 20 km wide belt of Middle to Late.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 99 of 132 Mineral Resources 2008 Explanatory Notes

Cambrian volcanic and non-volcanic sedimentary rocks that extend from Elliott Bay to Deloraine as an arc on the western and northern margin of the Precambrian Tyennan Region. Compositionally and texturally diverse lavas, syn-volcanic intrusions and syneruptive volcaniclastic units are the predominant volcanic facies associations of the Mt Read Volcanics. Virtually all facies were emplaced in a sub-aqueous environment below wave base. The lavas and intrusions are submarine, calc-alkaline primarily felsic bodies. Thick and extensive pumice-rich and crystal-rich units and in situ and resedimented hyaloclastite are typical of the volcaniclastic facies associations. The sedimentary facies are interbedded with the volcanic facies and include black mudstone and graded, bedded sandstone derived from volcanic and Precambrian metasedimentary basement sources.

The Rosebery base metal deposit is the largest massive sulphide deposit in the Mt Read Volcanics of western Tasmania. The deposit comprises of several polymetallic tabular lenses that vary from 0.1 – 5 Mt in size. The lenses are hosted in a felsic volcanic and sedimentary host sequence that dips 45o E and occur over a strike length of over 3000 m to at least 1500 m depth.

The lenses comprise massive and banded sphalerite, galena, pyrite, chalcopyrite and barite. Minor minerals include tetrahedrite, tennantite, arsenopyrite, pyrrhotite, hematite, magnetite, electrum, enargite, Fe bearing aikenite (a bismuthian sulfosalt), meneghinite,jordanite, bournonite and rutile. Chlorite, quartz, muscovite and several carbonates are the dominant gangue minerals with minor phases of spessartine, albite and coarse-grained barite. The bases of the lenses include assemblages rich in pyrite and chalcopyrite that may alternate with phyllosilicate bands in some lenses.

8.4 Assessment and Reporting Criteria

The following table provides a summary of important criteria related to the assessment and reporting of the Rosebery mineralisation. Note that as drilling was carried out in a number of phases from the start of current operations in 1936 to present day, techniques have varied over that time.

Criteria Status Sampling Techniques and Data Drilling • Current diamond drilling is NQ or BQ size for underground diamond drill techniques holes. • Current surface diamond drill-holes are HQ or NQ size. • Historical drill-holes are a mixture of sizes from AQ, LTK, BQ, NQ and PQ. • A variety of diamond drill types have been used over time. Drill Sample • Not quantified, generally high recoveries. Recovery Logging • Logged using Laptop computers and Corelog software from 1996 to present. • Prior to 1996 diamond drill holes were logged using Lotus spreadsheets or on paper.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 100 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Sub-sample • Geological samples are prepared as per Rosebery procedure RP-106- techniques 00003 on site at Rosebery mine. Samples prepared offsite are prepared and sample to the same specifications. preparation Quality of From 2005 Geological samples have been processed in the following manner: assay data and Dried laboratory • Crushed tests • Pulverised to 75 micron • Acid digest • Analysis of Pb, Zn, Cu, Ag, Fe by Atomic Absorption Spectrometry (AAS) • Au values are determined by fire assay • Sizing analysis is carried out on 1:20 pulps to ensure 90% passing 75 micron Verification of • QAQC program instigated in 1996 consisting of internal reference sampling and materials. assaying • Certified reference materials, duplicates and blanks introduced to QAQC program in 2008. • QAQC standards measured against a 3 standard deviation pass/fail, assessed via control charts for bias. • QAQC protocols defined by Rosebery procedures RP – 106-00020, RP- 106-00017 and RP-106-00023. Current QAQC protocols are under review. Location of • All current diamond holes are down-hole surveyed using a Reflex Ezi-shot Data Points tool at 30 m intervals. • Collar positions of underground drill-holes are picked up by Rosebery Mine Surveyors. • Collar positions of surface drill-holes are picked up by contract Surveyors using differential GPS. • Selected surface exploration holes have been down-hole surveyed using a gyro (parent holes only). Data spacing • The Rosebery mineral deposit is drilled on variable spacing dependent on and lens characteristics. Drill spacings are typically from 100x100m (inferred) distribution to 15x15m (measured, P lens).

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 101 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Orientation of • Drilling data is variably orientated with relation to the main orientation of data in the Rosebery mineral deposit. Underground and surface holes are drilled relation to to intersect mineralisation at as close to 90 degrees as possible. Drilling is geological undertaken from both footwall and hanging-wall positions to structure mineralisation. • Care need to be taken when planning and drilling diamond holes at Rosebery to account for regional foliation which can add to hole deviation. Estimating and Reporting of Mineral Resources Database • All Rosebery drill-hole data is stored in the Corelog2000 database (MS Integrity Access). • A Major database validation project was undertaken in march 2001, with a number of relatively minor errors found • Validation routines in the Corelog program check for overlapping sample, lithlogical and alteration information, as well as reject criteria such as logging information past EOH depth. Geological • Mineralisation at Rosebery consists of a series of massive sulphide lenses Interpretation hosted within felsic volcanic sandstones and siltstones (Host Sequence). • The Host package and sulphide lenses have an approximate 45 degree dip to the east, with some localised variation. • Geological modelling of the mineralisation follows geological and TMU boundaries. Each lens is interpreted separately as several mineralogical, metallurgical and physical differences occur between lenses relating to structure, alteration and primary mineralogy. Dimensions • The Rosebery mineral deposit extends from 400E-1800E, 2500N to - 1100N, 3400-1900RL (Rosebery Mine grid co-ordinates) and is currently open to the north, south and at depth. Individual lenses vary in size from a few hundred metres to up to 1000m along strike and/or down dip.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 102 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status TM Estimation and • Datamine software has been used to estimate Mineral Resources from Modelling 1999 to present. Techniques • For current Mineral Resource estimations a parent block size of 5 x 10 x 5 metres is used. Historically smaller block sizes have been used. • Various estimation techniques have been used historically at Rosebery including polygonal, nearest neighbour, inverse distance and ordinary kriging. Current block models use a mixture of inverse distance squared and ordinary kriging, guided by the wireframed geological and mineralogical domains. • Separate block models are created for each lens. • No metal top cuts are used. • 1m assay composites are used for all estimation work. • No dilution or recovery factors are taken into account during the estimation of Mineral Resources. • Maximum of two composites from any one hole. Moisture Tonnes have been calculated on a dry basis. Cut-off TMU Definition (Mineral Resource Cut-Off grade) parameters Rosebery has used a Total Metal Unit (TMU) as its cut-off grade for Mineral Resource modelling to reflect the multi-commodity nature of the Rosebery mineralisation for a number of years. It is an ex-smelter, recovered metal calculation. Variables in the TMU calculation are the grades of the individual metals, their net smelter recovery and the AUD/USD conversion rate. The current TMU cut off value in use for Mineral Resource estimation is AUD85. A minimum width criterion of 3m is most often used, however thinner intersections than this can be considered if continuity is high and grades sufficiently rich to enable the intersection to bulk out to meet the >3m at AUD85 TMU cut-off grade. TMU Calculation For Assays TMU (AUD) = ( Pb * b * c ) + ( Zn * d * e ) + ( Cu * f * g ) ( a * 10,000 ) ( a * 10,000 ) ( a * 10,000 ) + ( Ag * h * i * 0.03215 ) + ( Au * j * k * 0.03215 ) ( a * 100 ) ( a * 100 )

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 103 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status The following pricing assumptions were used in 2007/2008.

Metal Prices and Recoveries Parameters For 2007/2008

Code Description a AUD/USD Exchange Rate 0.85 b Pb Metal Price (USD/t ) 1,100 c Pb Recovery from feed to payable metal (%) 77.9 d Zn Metal Price ( USD/t ) 2,150 e Zn Recovery from feed to payable metal (%) 77.2 f Cu Metal Price ( USD/t ) 4,650 g Cu Recovery from feed to payable metal (%) 51.7 h Ag Metal Price ( USD/oz ) 12.5 i Ag Recovery from feed to payable metal (%) 69.1 j Au Metal Price ( USD/oz ) 750 k Au Recovery from feed to payable metal (%) 72.7 The metal price and exchange rate parameters used above are long term forecasts as issued by Zinifex (now OZ Minerals) Group Office in March 2008, and subject to change. Mining Factors Minimum mining width of 3 metres. or assumptions Metallurgical Recoveries are based on historical recoveries provided by the Rosebery factors or concentrator. assumptions Bulk Density Dry Bulk Density (DBD) Estimation Rosebery uses a formula to determine the dry bulk density (DBD), based on Lead, Zinc, Copper and Iron assays, and assuming a certain partition of the iron species between chalcopyrite and pyrite. The Rosebery DBD calculation had anecdotally been very representative of the mineralisation types. A study was conducted in August 1999 to examine the accuracy of the present algorithm compared to measured values. This study concluded the algorithm is accurately representing DBD for the Rosebery mineralisation. The formula is set out below: Gn% = Pb%/0.8658= 1.1550Pb Sp% = Zn%/0.6315789 = 1.5833Zn Cp% = Cu%/0.343 = 2.9155Cu Py% = (Fe%-(sp%*0.04)-(cp%*0.3043))/0.467

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 104 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Therefore Py% = 2.1413Fe%-0.1356Zn%-1.8997Cu%

Nsg = 100-(gn%+sp%+cp%+py%) Therefore Nsg = 100- 1.155Pb-1.5833Zn-2.9155Cu-2.1413Fe+0.01356Zn+1.8997Cu

SG = (gn%*0.075)+(sp%*0.04)+(cp%*0.042)+(py%*0.05)+(nsg*0.0265)

Gn%*0.075 = 0.866Pb% Sp%*0.04 = 0.0633Zn% Cp%*0.042 = 0.1224Cu% Py%*0.05 = 0.1071Fe%-0.0068Zn%-0.0950Cu% nsg*0.0265 = 2.65 –0.0306Pb% –0.0384Zn%-0.0269Cu%-0.0567Fe%

THEREFORE

SG =2.65+0.0560Pb%+0.0181Zn%+0.0005Cu%+0.0504Fe% Classification Mineral Resource Classifications

Inferred Mineral Resources Mineralisation of demonstrable continuity based on geological interpretation between more than one drill-hole. Drill-hole spacing nominally 100x100m or less. Indicated Mineral Resources Drill-hole spacing is a nominal 50m along strike and 50m up and down dip in the plane of mineralisation or less. There is some evidence of geological and grade continuity. Measured Mineral Resources The drill-hole spacing is a nominal 25m along strike and 25m up and down dip or less. Knowledge of the geology and grade distribution of the mineralisation is sufficiently high to allow detailed stope definition. In areas of a multi lens nature or complex geology closer drill-hole spacing may be required (For example P lens and remnant mining areas). Audits or An external review of the Mineral Resource estimation process was completed Reviews by Behre Dolbear Australia (BDA) in 2007.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 105 of 132 Mineral Resources 2008 Explanatory Notes

8.5 Compliance with the JORC Code Assessment Criteria

This Mineral Resource statement has been compiled in accordance with the guidelines defined in the Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves (The JORC Code, 2004 Edition).

The information in this report that relates to Mineral Resources is based on information compiled by Trevor Ellice, who is a Member of the Australasian Institute of Mining and Metallurgy and has sufficient experience which is relevant to the style of mineralisation and type of deposit under consideration and to the activity which they are undertaking to qualify as Competent Persons as defined in the 2004 Edition of the ‘Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves’. Trevor Ellice consents to the inclusion in the report of the matters based on their information in the form and context in which it appears.

Trevor Ellice is an employee of Oz Minerals Limited.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 106 of 132 Mineral Resources 2008 Explanatory Notes

9 AVEBURY MINE AREA RESOURCES - JANUARY 2008

9.1 Summary

This Mineral Resource estimate incorporates results received until December 2007.

Table 23 Avebury Mineral Resource as at 30 June 2008 reported above 0.4% Ni.

Tonnes Nickel Contained (Mt) Grade Metal (%Ni) Nickel (kt) Measured 2.37 1.03 24.4 Indicated 6.05 1.01 61.1 Inferred 9.76 0.88 85.9 Total 18.18 0.94 171.4 Above 0.4% Ni Cut off

Table 24 Avebury Mineral Resource as at 30 June 2008 reported above 0.7% Ni.

Tonnes Nickel Contained (Mt) Grade Metal (%Ni) Nickel (kt) Measured 2.22 1.06 23.5 Indicated 4.90 1.09 53.4 Inferred 6.88 0.99 68.1 Total 14.00 1.04 145.1 Above 0.7% Ni Cut off

9.2 Detail

All Mineral Resources quoted in this report were estimated from 3 dimensional block models created with SurpacTM Software. Mineral Resources are modelled on geological boundaries and/or a minimum 0.4% Ni cut off boundary which approximates the natural break between nickel mineralisation and background grades. A block cut off of 0.7% Ni was applied for the reserve calculation.

Ni, As and Co grades were interpolated using an ordinary kriging algorithm for all Mineral Resources with the exception of the East Avebury Inferred Resource. Geostatistical parameters for resource estimation were derived by Quantitative Geoscience consultants of Perth. The East Avebury Inferred Mineral Resource was estimated with an inverse distance squared algorithm.

Mineral Resource Estimation has been completed by Tim Callaghan (Allegiance Metals) and David Gibbons (Allegiance Metals) with consulting advice and assistance of Scott Jackson, (Quantitative Geoscience). Geostatistical analysis and variogram modelling was completed by Scott Jackson using Isatis Software in December 2007.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 107 of 132 Mineral Resources 2008 Explanatory Notes

9.3 Setting

The Avebury Nickel Sulphide Mine is located 10km west of Zeehan on the west coast of Tasmania. This estimate covers all Mineral Resources contained on ML3M/2003 and ML6M/2007, held by Allegiance Metals, a wholly owned subsidiary of OZ Minerals.

The Avebury Nickel Sulphide deposit is hosted in moderately to steeply dipping Cambrian ultramafic intrusives belonging to the McIvor Hill Mafic-Ultramafic Complex.

The whole sequence has undergone moderate contact metamorphism to hornfels accompanied by mild to strong metasomatism during the intrusion of the Heemskirk Granite at the end of the Devonian Tabberabberan Orogeny. Variable metasomatism of the ultramafics has formed two distinctly different gangue mineral assemblages; a serpentinite-magnetite gangue or an intensely metasomatised tremolite-diopside-magnetite gangue. Most of the nickel sulphide mineralisation is located within the ultramafic immediately adjacent to its margins. Nickel grades diminish toward the interior of the intrusions. Mineralisation is dominated by a pentlandite-pyrrhotite-magnetite assemblage with much lesser millerite, gersdorfite and niccolite.

Mineralised widths vary from 4 to 40m and average around 10m true width. Mineralised lenses are generally around 50-600m in length and can extend over 400m down dip.

9.4 Assessment and Reporting Criteria

The following table provides a summary of important criteria related to the assessment and reporting of the Avebury Mineral Resources.

Criteria Status Sampling Techniques and Data

Drilling • NQ diamond core or equivalent LTK60. Techniques • 336 diamond holes for 88,000m (126 surface the remainder underground).

Sample recovery • Excellent. Generally 100%, with minor losses in broken ground. There is no relationship between core loss and mineralisation.

Logging • Geological and geotechnical logs on excel spreadsheets by experienced geologists. • All core is stored in one of the two Zeehan core sheds. • Drill logs are downloaded directly into site Access Database

Sub -Sample • Half core split on 1m samples while respecting geological contacts. preparation • Core is split by diamond saw, bagged, numbered and dispatched to Assay laboratories. • Samples are generally 2kg in weight. • Crush, mill and homogenise entire sample to <0.75 micron.

Sample • Pre 2005: 4 acid digest and analysis of Ni, As, Co, S by ICP_AES at

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 108 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status preparation Analabs laboratories, Townsville. • Post 2005: Pressed powder XRF analyses for Ni, As, Co, S, FeO, MgO at Burnie Research Laboratories.

Assay QA/QC • Submission of internal standards and pulp duplicates with every batch of samples. • Approximately 1 in every 10 submissions sent for independent laboratory analysis (Amdel laboratories, Adelaide) by ICP_AES.

Location of Data • All hole collar surveys by licensed surveyor. • All coordinates in Mine Grid Plane Projection (close approximation of AGD66). • Strong local magnetic fields associated with Avebury mineral deposit reduce the effectiveness of conventional down-hole survey tools. • Post 2005 surface DDH gyroscopically surveyed. • Pre 2005 surface holes surveyed by Maxibor or had azimuth corrected from nearby Maxibor holes. • Down-hole dip taken from digital survey tool or Eastman single shot cameras. • Underground DDH Azimuth taken as collar azimuth east of 345350. o • Underground DDH Azimuth taken as collar azimuth +1 /50m west of 345350.

Data Spacing and • Drill spacing approximately 25 x 20 or 20 x 50 for better drilled areas distribution (Measured and Indicated). • Spacing 100 x 100 or 100 x 50 for Inferred Mineral Resources. • All holes drilled North-South or South-North, perpendicular to the strike of mineralisation. Estimating and Reporting of Mineral Resources

Database Integrity • All data stored in customised access database. • All logging entered into excel logging sheets and downloaded into database. • Assay data loaded from excel spreadsheets directly into database. • Data integrity validated with Surpac Software for EOH depth and sample overlaps. • Manual check by plotting sections and plans. • Drill hole A007 has been removed from the database due to inaccurate survey results.

Geological • Mineralisation is hosted in Middle Cambrian ultramafic bodies Interpretation intruding Cambrian volcaniclastic sediments. • Both host volcaniclastic and ultramafic intrusions are steeply north dipping in an overturned south facing sequence.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 109 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

• The stratigraphy and intrusions broadly strike East-West. • Devonian Granite intrusion has strongly hornfelsed and locally metasomatised the host sequence. • Gangue mineralogy is either black serpentinite-magnetite or a pale grey-green diopside-tremolite-magnetite. • Mineralisation consists of coarse disseminated and stringer pentlandite with minor pyrrhotite. • Nickel arsenides (Niccolite, Maucherite, Gersdorfite), although sparse are a penalty element and occur in discreet cross cutting zones. • Mineral Resource estimation was made with Surpac Software. Separate Nickel, Arsenic and Ultramafic domains were wire frame modelled using north-south cross sections, respecting geological contacts and down hole geochemical data. • Nickel domains are delineated on a 0.4% Ni cut off which is the natural break between background silicate Nickel and Nickel Sulphides. Coarse pentlandite mineralisation is visible above 0.4% Ni. • Separate wire frames were modelled for high arsenic (>300ppm) domains in the Avebury Model.

Dimensions • Mineralisation extends between: 353700mE to 355900mE 5357100mN to 5357750mN 1550mRL to 2200mRL

Estimation and • Two block models were used in the estimation: Modelling Avebury Block Model - avebury108.mdl techniques East Avebury Block Model - east_avebury108.mdl The Avebury Model covers the current mine area and is the basis for the Ore Reserves. Block model attributes were interpolated using an ordinary kriging algorithm. The East Avebury model includes Inferred Mineral Resources located east of the Avebury Mine. Block model attributes were interpolated using inverse distance squared weighting as limited drilling information negated the need for more sophisticated estimation techniques. • Parent block size was set to 10 x 10 x 10m with sub blocks to 2.5m in the x and z orientation and 1.25 in the y orientation. • 1m assay composites were used for the estimation. • Fractional composites are not used in the estimation if they are less than half the composite length. • Geostatistical analysis and variogram parameters were supplied by Quantitative Geoscience (QG). Tables of search ellipse and variogram parameters are tabulated at the end of this document.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 110 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

• Nickel domains were used to estimate Co, S, MgO and bulk density in the Avebury Model. Background Arsenic was interpolated for the entire ultramafic in the Avebury Model using a top cut of 1000ppm As. High Arsenic domains in the Avebury Model were interpolated without top cutting. • Nickel Domains only were used in the estimation of all parameters in the East Avebury model. Arsenic was top cut to 1000ppm.

Moisture • Tonnes have been calculated on a dry basis.

Cut-off • Global in-situ mineralisation tonnes and grade have been estimated parameters. from within the >0.4% Ni mineralisation shape. • Mineral Resources have been reported at a 0.4% Ni and 0.7% Ni block cut-off.

Mining Factors or • No mining assumptions have been made. assumptions

Metallurgical • Metallurgical test work has been completed for the Measured and factors and Indicated areas of the Mineral Resource and selected areas of the assumptions. Inferred Mineral Resources. Test work includes standard variability comminution, grinding and float tests and the treatment of bulk samples from selected mineralisation types.

Bulk Density • Bulk density measurements are undertaken by the weight in air/weight in water technique. • The Avebury Nickel Sulphides are hosted by two distinct gangue mineralogies with distinct SG properties. These are either a serpentinite host (SERP) with an SG of 2.8, or an ultramafic skarn host (SKSP) with an SG of 3.2. Tonnage calculations for the block model were made using an indicator method where the composited rock types were assigned either a 0 for SERP or a 1 for SKSP per 1m interval. A ratio of the two rock types for each block was interpolated from these data points using inverse distance squared weighting. The blocks were then assigned an SG via the following formula: SG(block) = 2.8 + SKSP/SERP(block) x (3.2 – 2.8) Where SKSP/SERP(block) = the interpolated SKSP/SERP ratio for each block. Every block is then assigned an SG value between 2.8 and 3.2 depending on the interpolated ratio of SERP to SKSP. • Background bulk density outside of mineralised domains and for waste rock was assigned a value of 3.

Classification • Measured Mineral Resource Distance to nearest sample <30m

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 111 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Drill spacing <50 x 20m Accessed by mine level development • Indicated Mineral Resource Distance to nearest sample <30m Drill spacing <60 x 40m • Inferred Mineral Resource The remainder of the modelled domains.

Validation • Global statistical analysis between estimated blocks and input data. • Visual checks of block grades and drill hole data in plan and section • Manual sectional Mineral Resource estimate.

Audits and • Full audit of 2005 Resource Estimate by AMC Consultants. reviews • Reviews of Mineral Resource estimate by AMC Consultants for 2007 and 2008.

• Review by Behre Dolbear Australia (BDA) 2007.

9.5 Competent Person Statement

The Mineral Resource statement has been compiled in accordance with the guidelines defined in the Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves (The JORC Code, 2004 Edition).

The information in this report that relates to Mineral Resources is based on information compiled by Tim Callaghan, who is a Member of the Australasian Institute of Mining and Metallurgy and has sufficient experience which is relevant to the style of mineralisation and type of deposit under consideration and to the activity which they are undertaking to qualify as Competent Persons as defined in the 2004 Edition of the ‘Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves’. Tim Callaghan consents to the inclusion in the report of the matters based on their information in the form and context in which it appears.

Tim Callaghan is a full time employee of Oz Minerals Limited.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 112 of 132 Mineral Resources 2008 Explanatory Notes

10 HIGH LAKE MINERAL RESOURCE STATEMENT - 30 JUNE 2008

10.1 Summary

The High Lake Mineral Resource estimate is as determined by Wardrop Engineering in June 2006 as shown in the table below.

Table 25 High Lake Mineral Resource as at 30 June 2008

Category Tonnes Zn Cu Pb Ag Au (kt) (%) (%) (%) (g/t) (g/t) Measured 0 Indicated 17,200 3.35 2.25 0.31 70 0.95 Inferred 42 2.38 0.49 0.44 122 0.21 Total Measured and Indicated 17,200 3.36 2.25 0.31 70 0.95 Total Measured, Indicated and Inferred 17,242 3.35 2.25 0.31 70 0.95

The Mineral Resource is reported at a 2% copper equivalent cut-off based on the parameters used in 2006.

10.2 Introduction

The High Lake Mineral Resource Statement covers the copper-zinc-silver-lead-gold Mineral Resources for the High Lake deposits located approximately 550 kilometres north-northeast of Yellowknife in the Nunavut Territory of Canada. The known Mineral Resources are contained within three zones, the AB, D and West Zones.

The AB and D zones were discovered and evaluated by Kennecott in the 1950s. The property then remained dormant until Kennecott and Aber Resources re-started exploration in 1991. Wolfden Resources acquired the property in 2001 and discovered the main West Zone mineralisation, 1.3 kilometres west of the D zone in 2003. Zinifex acquired Wolfden in 2007.

This Statement is based on a technical report entitled “Preliminary Economic Assessment of the High Lake Property” prepared by Wardop Engineering of Toronto Canada in June 2006 and for the then owners of High Lake, Wolfden Resources Inc. The report was prepared in compliance with the Canadian reporting standard National Instrument (NI) 43-101 and released into the public domain on “SEDAR”, the company disclosure site in Canada. NI43-101 requires the reporting of Mineral Resources and reserves to conform to the standards set by the Canadian Institute of Mining, Metallurgy and Petroleum (“CIM”), a recognised overseas professional organisation (“ROPO”) under the JORC Code (2004 edition). Other reports supporting this latest report are “High Lake Property Resource Estimate” by G.H. Wahl dated 8th January 2005 and “Qualifying Report on High Lake Property Nunavut” by G.A Harron dated 14th April 2003.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 113 of 132 Mineral Resources 2008 Explanatory Notes

The Mineral Resource estimates were based on the historical geological database which included all the current drilling up to 2005. The geological interpretation, geological modelling, block model construction and grade estimation were completed by G Wahl and Associates for the AB and D zones using SurpacTM software and by G Wahl and Wardrop Engineering for the West Zone using DatamineTM software.

A desktop review of the Canadian Mineral Resources including High Lake was conducted by an independent consultant for Zinifex in 2008 and found the Mineral Resource estimate to be generally reasonable and appropriate.

10.3 Setting

Figure 13 Location of High Lake

The High Lake copper-zinc project is located approximately 550 kilometres north-northeast of Yellowknife in the Nunavut Territory of Canada. The known Mineral Resources are contained within three zones, the AB, D and West Zones.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 114 of 132 Mineral Resources 2008 Explanatory Notes

The High Lake VMS deposits are hosted within the High Lake greenstone belt in the northern part of the Slave structural province. Basement gneisses about 3.96-3.00 Ga are overlain by sedimentary and volcanic rocks (2.69 Ga). Three phases of regional deformation are evident and the province is intruded by batholiths ranging between 2.63-2.58 Ga. Generally, the supracrustal rocks trend north to northeast. The High Lake greenstone belt extends 140 km from the Coronation Gulf coast southward and ranges from 5 to 30 km in width. The belt is characterised by a higher proportion of felsics over mafic volcanics when compared to other Slave province greenstone belts. The belt exhibits greenschist grade metamorphism and is intruded by syn- volcanic mafic to intermediate plutons and post tectonic felsic intrusions. Proterozoic age (1.27 Ga) northwesterly trending diabase dykes, part of the Mackenzie swarm, intrude the supracrustal sequence.

The High Lake stratiform massive sulphide deposits are hosted by felsic volcanic rocks near or adjacent to a granodiorite intrusion and are considered volcanogenic in origin (VHMS type) with well developed hydrothermal alteration and stringer zones.

Figure 14 AB, D and West Zone Locations

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 115 of 132 Mineral Resources 2008 Explanatory Notes

10.4 Assessment and Reporting Criteria

The following table provides a summary of important criteria related to the assessment and reporting of the High Lake mineralisation. Note that drilling was carried out in a number of phases from 1956 to 2006 and techniques varied over that time.

Criteria Status Sampling Techniques and Data Drilling • 51 AX sized diamond drill holes for 7,050 metres (1956-57) techniques • 61 NQ sized diamond holes for 10,693 metres (1992-93) • 173 NQ sized diamond holes for 55,581 metres (2001-05)

Drill Sample • Unknown for the two earlier drilling campaigns Recovery • 97% for the 2001-2005 drilling

Logging • Logged on paper in the earlier campaigns • Logged on Excel sheets for the 2001-2005 drilling

Sub-sample • No details from 1956-57 drilling. techniques • For later drilling campaigns core was sawn in half with one half bagged and sample and sent for assay. Details of the sample preparation are included below. preparation • Sampling intervals ranged from 0.25-2.5 metres with most being around one metre. Sampling honoured the logged geological boundaries.

Quality of • Early assays were done at Kennecott’s laboratory in Salt Lake City but no assay data and details available. laboratory • For the 1991-92 program base metals were analysed by aqua regia tests digestion of a 0.5g sample and ICP-AAS determination. Gold and silver were determined by fire assay. • The 1993 assay determinations involved aqua regia digestion followed by determination of Cu, Pb, Zn and Ag by flame AAS. Gold assays were determined by FA-AAS using a 30g sample. For samples containing >10g/t Au, assays were re-run using a standard fire assay technique. • Acme Analytical Laboratories in Vancouver were use in 1991-92 and Chemex Laboratories of Vancouver in 1993. • From 2001, assaying was carried out by three accredited commercial laboratories: ALS Chemex Ltd of Vancouver in 2001, Accurassay Laboratories of Thunder Bay, Ontario in 2002, and Global Discovery Laboratories (a division of Teck Cominco) in subsequent years. • In the 2001 drill campaign samples were crushed to 70% passing -6mm. A riffle splitter was used to cut a <250 gram sample which was then pulverized to 85% passing 75 micron. Silica sand washes were used between samples. Assaying of Cu, Zn, Pb and Ag were completed using aqua regia digestion followed by AAS. Gold assays were based on a 20gram pulp using fire assay and an AAS finish. Insertion of control

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 116 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status samples was reported to have occurred. • In 2002 samples were crushed to -8 mesh and a 400 gram sample was extracted using a riffle splitter. Samples were pulverised to 95% passing - 150 mesh. Silica sand washes were used between samples to minimize cross sample contamination. Samples were assayed for Cu, Pb, Zn and Ag using aqua regia digestion and AAS. Gold assays were based on a 20 gram pulp using fire assay and an AAS finish. Insertion of control samples was reported to have occurred. • From 2003 half core was dried and then crushed in two stages to -6mm and then to -2mm. The sample was then split in a riffle splitter to obtain a sub-split of 250-300 grams. The sub-split was then pulverized in a ring and puck mill to produce a final product that was 95% minus 200 mesh (minus 75 microns). The ring and puck mills were cleaned using pure quartz sand before each batch and were cleaned between samples if there was any apparent residue in the mills. The rejects were placed in new plastic bags, rolled to remove all excess air and taped up. These were then boxed up at 8-10 samples per box and stored at room temperature in dry a environment should any additional pulps be required for metallurgical test work. Samples were assayed for Cu, Pb, Zn and Ag using aqua regia digestion and AAS. Gold assays were based on a 30 gram pulp using fire assay and an AAS finish.

Verification of • Re-sampling of old drill core from the 1956-57 drilling was carried out by sampling and Wolfden and G Wahl in 2005 and did not identify any significant issues assaying with the assay database. • Check sampling of pulps from the 1992-93 drilling was also carried out with no significant issues identified.

Location of • All diamond holes were surveyed in local grid coordinates and details Data Points recorded on the drill logs. • Survey method unknown. • A check of all drill-hole collars was conducted by G Wahl in 2005. A number of errors were noted in the Kennecott and Aber drilling and corrected. A persistent error of several metres was noted for the West Zone drilling and a recommendation made to resurvey holes in this area which was subsequently done. • Only start and end down-hole acid surveys were generally recorded for holes for the Kennecott and Aber drilling. Most of these holes were less than 200 metres in length, the longest being 374 metres. • Later holes had down-hole Easyshot surveys at 50-100 metre intervals.

Data spacing • The AB and D zones were drilled off on 20-40 metre spaced sections. The and West Zone was drilled on 25-50 metre spacing down to around 800 distribution metres below surface.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 117 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Orientation of • Most holes drilled at right angles (grid west-east) to the stratigraphy. data in • Intersection angles vary but typically 30-40 degrees through the lenses. relation to geological structure

Estimating and Reporting of Mineral Resources Database • For the Kennecott and Aber drilling programs all data was recorded on Integrity paper. • Wolfden logged on Excel spreadsheets • All data was subsequently transferred to an electronic database • The procedures for checking the integrity of the database are unknown.

Geological • Mineralisation is stratiform within a felsic volcanic sequence. Interpretation • Granodiorite, granitoids and pegmatite locally intrude the sequence and are in turn cut by a suite of diabase dykes. • The AB zone is made up of a large stringer zone (the “A” zone) stratigraphically below a massive sulphide lens (the “B” zone). A diabase dyke separates the two zones. 12 separate lenses were modelled. • The D zone is made up of four banded polymetallic sulphide lenses, striking at 30 degrees and dipping steeply to the west-northwest. • The West zone is made up of three lenses with the bulk of the mineralisation contained in one lens. The lenses dip sub-vertically to the east. • For the AB and D zones geology codes and a 1% Cu equivalent grade were used as guide to build the wireframes using SurpacTM software. Interpretations were digitised on 20-40 metres spaced sections and snapped to drill-holes. A minimum width of 3 metres was used. Twelve triangulated solids were interpreted for the AB zone and four for the D zone. • For the West Zone geology codes and a 1% Cu equivalent grade were used as guide to build the wireframes using DatamineTM software. Interpretations were digitised on 25 metre spaced sections and used to build the triangulated solids. Three solids were interpreted for the West Zone.

Dimensions • Twelve solids were created for the AB zone, the largest of which had dimensions of 150m in length, 360m down dip and up to 80m in width. • The D zone is made up of four lenses, the largest of which had dimensions of 150m in length, 320m down dip and up to 35m in width. • The West zone is made up of three lenses, the largest of which had a strike of 275m, extended to 900m down dip (and is still open) and up to 40m thick.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 118 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

TM Estimation and • Datamine software was used. Modelling • For Mineral Resource estimation of the AB and D zones a block size of 10 Techniques x 5 x 5 metres was selected by Wahl covering all the known mineralisation. The area covered by the block model, express as UTM NAD 83 grid coordinates are set out below.

UTM NAD 83 (metres) Coordinates Coordinate Minimum Maximum Northing 7472700 7477400 Easting 505900 506950 Elevation -200 380 One metre composites were used. Block grades were estimated by the inverse distance squared method.

For the West Zone a block size of 2.5 x 10 x 10 was selected. The area covered by the block model expressed as UTM NAD 83 coordinates are:

UTM NAD 83 Coordinates Coordinate Minimum Maximum Northing 7472000 7473000 Easting 504400 505400 Elevation -600 400 One metre composites were used. Block grades were estimated by the inverse distance squared method.

• Top cuts were applied to high grades for precious metals only as shown below. Deposit Au g/t Ag g/t West Zone 9.2 300 AB 3.7 120 D 2.7 190 • No dilution or recovery factors were taken into account.

Moisture • Tonnes have been calculated on a dry basis

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 119 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Cut-off • A 2% CuEq cut-off was applied to define the mineralisation parameters • The CuEq formula used was: CuEq% = Cu% + (Zn%*(Zn recovery*Zn price)/(Cu recovery*Cu price) + (Au g/t* (Au recovery*Au price)/31.103)/(Cu recovery*22.046*Cu price)

+ (Ag g/t* (Ag recovery*Ag price)/31.103)/(Cu recovery*22.046*Cu price)

The following assumptions were used by Wahl and Wardrop: Metal Recoveries Price Zinc 70% USD0.48/lb Copper 85% USD1.00/lb Silver 75% USD5.00/oz Gold 75% USD375/oz Mining Factors • A minimum mining width of 3 metres is used. or assumptions

Metallurgical • Recoveries supplied by consultant metallurgist John Knapp in 2005 based factors or on test work. assumptions

Bulk Density • Specific gravity applied to the block models were based on averages determined from physical measurements for mineralisation and waste. Details are summarised below. Zone No Samples Mineralisation Waste Average SG Average SG West Zone 77 3.69 2.84 AB 52 3.22 2.84 D 52 2.96 2.84 Classification • The classification of an Indicated Mineral Resource is as follows: West Zone

• Percentage of block must lie within the interpreted mineralized wireframe. • Minimum of three composites, and a maximum of 12 are used to estimate a block. • Distance to composites is less than or equal to 75 metres. • Search parameters are shown below Search Ranges (m) Metal X Y Z

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 120 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Ag 150 150 37.5 Au 260 260 26 Cu 260 260 26 Pb 150 150 37.5 Zn 250 250 25 Zone Z Y X 100 30 80 0 200 19 74 0 300 16 73 0

• AB & D Zones - Percentage of block must lie within the interpreted mineralized wireframe. - Minimum of three composites and a maximum of 12 are used to estimate a block. - Search parameters as shown below: Solid Major Axis Major/minor Major Axis Major Axis Range Axis Range Azi Dip Ratio AB_400 100 10:1 260 -90 AB_500 100 5:1 270 -64 AB_600 100 10:1 270 -55 AB_700 100 10:1 270 -85 AB_800 100 10:1 278 -70 AB_900 100 2:1 270 -68 AB_1000 100 10:1 270 -63 AB_1100 100 10:1 270 -80 AB_1200 100 10:1 252 -80 AB_1300 120 10:1 258 -65 AB_1400 100 10:1 265 -60 AB_1500 150 3:1 250 -68 D_1600 150 10:1 265 -58 D_1700 120 6:1 295 -75 D_1800 120 3:1 270 -78 D_1900 140 7:1 270 -90

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 121 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

• The remainder of the blocks within the interpreted three-dimensional wireframe was classified as being Inferred Mineral Resources.

Audits or • A desktop review of the Canadian Mineral Resources including High Lake Reviews was conducted by Behre Dolbear Australia (BDA) for Zinifex in 2008 and found the Mineral Resource estimate to be generally reasonable and appropriate.

10.5 Competent Person Statement

The information in this report that relates to Mineral Resources is based on information compiled by the below listed competent person, who is a Members of a ‘Recognised Overseas Professional Organisation’ (‘ROPO’) and has sufficient experience which is relevant to the style of mineralisation and type of deposit under consideration and to the activity undertaken to qualify as Competent Persons as defined in the 2004 Edition of the ‘Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves’. The Competent Person consents to the inclusion in the report of the matters based on their information in the form and context in which it appears.

George Wahl is a member of the Association of Professional Geoscientists of Ontario and of the Association of Professional Engineers, Geologists and Geophysicists of the Northwest Territories and is a “Qualified Person” as defined in National Instrument 43-101 (NI 43-101) Standards of Disclosure for Mineral Projects of the Canadian Securities Administrators.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 122 of 132 Mineral Resources 2008 Explanatory Notes

11 IZOK LAKE MINERAL RESOURCE STATEMENT - 30 JUNE 2008

11.1 Summary

The Izok Lake Mineral Resource estimate is as determined by Wardrop Engineering in June 2006 as shown in the table below.

Table 26 Izok Lake Mineral Resource as at 30 June 2008

Category Tonnes Zn (%) Cu (%) Pb (%) Ag (kt) (g/t) Measured 0 Indicated 14,401 12.94 2.51 1.28 71 Inferred 369 6.40 3.79 0.27 54 Total Measured and Indicated 14,401 12.94 2.51 1.28 71 Total Measured, Indicated and Inferred 14,770 12.77 2.54 1.25 71

The Mineral Resource is reported at a 2% zinc equivalent cut-off based on the parameters used in 2006.

11.2 Introduction

The Izok Lake Mineral Resource Statement covers the copper-zinc-silver-lead Mineral Resources for the Izok Lake project located approximately 360 kilometres north of Yellowknife in the Nunavut Territory of Canada. The known Mineral Resources are contained within four lenses, the Central, Northwest, North and Inukshuk. The project is wholly owned by OZ Minerals.

The deposit was discovered and evaluated by Texasgulf in the 1970s. Falconbridge acquired the property in 1985 through its purchase of Kidd Creek Mines (formerly Texasgulf) but did no work. Minnova purchased the property in 1991 and subsequently sold a 40% stake to Metall. Metall and Minnova merged in 1993 and re-named Inmet in 1995. Further evaluation lead to a feasibility study being completed in 1994. Wolfden Resources purchased the property from Inmet in 2006. Wolfden was acquired by Zinifex in 2007.

This Statement is based on a technical report entitled “Preliminary Economic Assessment of the Izok Lake Property” prepared by Wardop Engineering of Toronto Canada in June 2006 for the then owners of Izok Lake Wolfden Resources Inc. The report was prepared in compliance with the Canadian reporting standard National Instrument (NI) 43-101 and released into the public domain on “SEDAR”, the company disclosure site in Canada. NI43-101 requires the reporting of Mineral Resources and reserves to conform to the standards set by the Canadian Institute of Mining, Metallurgy and Petroleum (“CIM”), a recognised overseas professional organisation (“ROPO”) under the JORC Code (2004 edition).

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 123 of 132 Mineral Resources 2008 Explanatory Notes

The Mineral Resource estimate was based on historical geological database which included all the current drilling up 2006. The geological interpretation, geological modelling, block model construction and grade estimation were completed by Wardrop Engineering using MineSight software.

A desktop review of the Canadian Mineral Resources including Izok Lake was conducted by an independent consultant for Zinifex in 2008 and found the Mineral Resource estimate to be generally reasonable and appropriate.

11.3 Setting

The Izok Lake copper-zinc project is located approximately 360 kilometres north of Yellowknife in the Nunavut Territory of Canada. The known Mineral Resources are contained within four lenses, the Central, Northwest, North and Inukshuk.

Figure 15 Izok Lake Location

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 124 of 132 Mineral Resources 2008 Explanatory Notes

The Slave Structural Province is predominantly underlain by Archean volcano-sedimentary rocks of the Yellowknife Supergroup and is locally divided into two geological formations, the lower Point Lake and the Contwoyto. The lower Point Lake Formation consists of a series of felsic to mafic tuffs, flows and metasediments with some calcareous rocks. The Contwoyto Formation is applied to a unit of metamorphic equivalents of greywackes and mudstones of turbidite origin, which also includes the presence of scattered discontinuous bands and lenses of silicate, sulphide or oxide facies iron formation. These formations were intruded by a late Archean granitic bodies and crosscut by north to northwest trending diabase dikes that belong to the Helikian-aged Mackenzie Swarm. Samples of the Izok Lake area rhyolite were dated using U-Pb zircon geochronology at 2623 ± 20 Ma.

The Izok Lake massive sulphide deposits are hosted by rhyolite within or near the top of the Point Lake Formation and are considered volcanogenic in origin (VHMS type). Strong hydrothermal alteration is evident around the deposits on which is superimposed high-grade (Amphibolite Facies) regional and contact metamorphism. Common metamorphic minerals include muscovite, biotite, sillimanite, cordierite and chlorite with local intense silicification. As a result the sulphides are coarse grained and have excellent metallurgical properties.

11.4 Assessment and Reporting Criteria

The following table provides a summary of important criteria related to the assessment and reporting of the Izok Lake mineralisation. Note that drilling was carried out in a number of phases from 1975 to 1995 and techniques varied over that time.

Criteria Status Sampling Techniques and Data

Drilling • 146 AQ sized diamond drill holes for 15,705 metres (1975-77) techniques • 223 NQ sized diamond holes for 45,932 metres (1992-95)

Drill Sample • Not quantified but reported to be very good Recovery

Logging • Logged on paper

Sub-sample • No details from 1975-77 drilling. techniques • For 1992-95 drilling, core sawn in half with half sample crushed to 6mm on and sample site. 250g sample riffled split out and sent for assay with rejects stored in preparation Yellowknife

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 125 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Quality of • No details for 1975-77 drilling. assay data and • For 1992-95 drilling samples were ring-pulverised to –150 mesh and a 0.5g laboratory sample taken to dryness with 15ml concentrated hydrochloric acid (HCl) tests and 5ml concentrated nitric acid (HNO3). The sample was then re-

dissolved in 5ml HNO3 and diluted to 100ml with deionised water for the copper, zinc and lead analyses. Atomic Absorption Spectrometry (AAS) was used to determine the elements.

• For silver a 2g sub sample was digested with 7.5ml HCl and 2.5ml HNO3 for one hour in a covered beaker. The sample was subsequently diluted to 100ml with 1:1 HCl and analysed by AAS. • For gold a 10g sub-sample was fused, cupelled and the subsequent dore bead parted with dilute nitric acid solution. The gold acquired through this process was rinsed with deionised water, annealed and weighed on a microbalance scale. • Sample repeats were conducted for every 20 samples and standards routinely inserted.

Verification of • Check sampling was conducted by Metall on some 1992-93 and 1975-77 sampling and drilling samples. No material discrepancies were identified assaying

Location of • All diamond holes were surveyed in local grid coordinates and details Data Points recorded on the drill logs. • Survey method unknown. • Collar positions of 28 holes checked by Wardrop in 2006 using GPS. Minor discrepancies recorded. • Only start and end down hole acid surveys were generally recorded for holes up to HEN341. Most of these holes were less than 200 metres in length. • Later holes had down hole surveys at various intervals, typically 50 metres or 100 metres.

Data spacing • The Central zone is drilled on 30 metre east-west section spacing (15 and metres on section), the Northwest and North lenses on 30 metre section distribution spacing and the Inukshuk lens on 50 metre section spacing.

Orientation of • The Central, Northwest and North zones are generally flat lying or gently data in dipping and were drilled mainly with vertically drill holes. relation to • The Inukshuk zone had a steeper dip and plunged 45 degrees to the east geological and was drilled with angled drill holes. structure

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 126 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status Estimating and Reporting of Mineral Resources

Database • All data was recorded on paper initially and was subsequently entered into Integrity an Access database • The procedures for checking the integrity of the database are unknown

Geological • Mineralisation is stratabound, being hosted within rhyolites near the top of Interpretation the Point Lake Formation, a sequence of intermediate to felsic volcanic rocks and lesser sedimentary rocks. • Gabbro, granitoids and pegmatite locally intrude the sequence and are in turn cut by a suite of diabase dykes. • The rocks are folded into a broad antiform with limbs striking northeast and southeast and dipping steeply to the SE and NE respectively. • Mineralisation is contained within four lenses. The North, Northwest and Central lenses are gently folded with shallow dips. The Inukshuk lens has a distinct plunge of around 45 degrees to the east. • Geological modelling of the mineralisation followed the scheme adopted by Metall which divided the mineralisation into three classes defined by mineralisation type.

Classes by mineralisation Type:

Class Mineralisation Type Copper - Pyrite-chalcopyrite (± pyrrhotite, sphalerite) - Chalcopyrite-pyrrhotite (± pyrite) Polymetallic - Polymetallic - Sphalerite-galena Zinc - Pyrite-sphalerite - Sphalerite-pyrite (± chalcopyrite) • The Central and Northwest lenses were interpreted based on the three classes. Inukshuk and North lenses used a single class type. • The final mineralised interpretation was conducted on 30-metre mine grid sections for each class and based on an approximate ZnEq cut-off of 2%. The creation of wireframes varied depending on the lens. For the North and Inukshuk lenses triangulated wireframes were created from the sectional interpretations. However for the Central and Northwest lenses the wireframe solids were created by extruding the 30 metre section outlines 15 metres (half distance) to the next section. Wardrop stated that this was done because of the complicated outlines of the mineralisation classes which made “normal’ triangulated wireframing difficult.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 127 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Dimensions • The Central, Northwest and North lenses extend from 30100E-31000E, 29750N-30250N, 10250-10400RL and are largely closed off. • The Inukshuk lens extends from 31150E-31400E, 29700N-29800N, 10100- 10400RL and is open down plunge to the east. TM Estimation and • Minesight software was used. Modelling • For Mineral Resource estimation a block size of 5 x 5 x 5 metres was Techniques selected by Wardrop covering all the known mineralisation. The area covered by the block model, express as mine grid coordinates are shown below. Izok Block Model Coordinates

Mine Grid Coordinate(metres) Coordinate Minimum Maximum X 30,000 31,500 Y 29,500 30,500 Z 10,100 10450 • For the Northwest and Central lenses grade interpolation was done by ordinary kriging guided by the wireframed geological and mineralogical domains. For the North and Inukshuk lenses inverse distance squared was used due to the more limited drilling information. • Top cuts were applied to high grades for all metals as shown below. Metal Top Cuts

Class Cu % Zn % Pb % Ag % Copper 20 35 2.5 300 Polymetallic 13 47 14 400 Zinc 11 42 10 300 • 2.5 metre assay composites were used. • Commonly the compositing of down-hole assays results in a fractional length in the last sample of a given hole. These fractional composites are not used in the Mineral Resource estimation if they are less than half the composite length. • No dilution or recovery factors were taken into account. • Maximum of two composites from any one hole.

Moisture • Tonnes have been calculated on a dry basis

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 128 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Cut-off • A 2% ZnEq cut-off was applied to define the mineralisation parameters • The ZnEq formula used was: ZnEq% = Zn% + (Cu%*(Cu recovery*Cu price)/(Zn recovery*Zn price)) + (Pb%*(Pb recovery*Pb price)/(Zn recovery*Zn price)) + (Ag g/t* (Ag recovery*Ag price)/31.103)/(Zn recovery*22.046*Zn price))

The following assumptions were used by Wardrop.

Metal Prices and Recoveries Metal Recoveries Price Zinc 94.5% USD0.60/lb Copper 90.0% USD1.20/lb Lead 80.0% USD0.45/lb Silver 70.0% USD7.50/oz Mining Factors • No mining assumptions have been made or assumptions

Metallurgical • Recoveries based on test work carried out in 1992-1994 as part of the Izok factors or feasibility study. assumptions

Bulk Density • Initially, specific gravity measurements were determined by Minnova in 1992 on 292 samples using the Archimedes method. • A calculated specific gravity method was subsequently developed in lieu of physical measurements based on visual estimates of sulphides and gangue and assays. The variables for the SG formula are in the table below. SG Formula Variables and Inputs

Variable Calculation Description GNE Estimated gangue % (assumed correct) CU Actual copper assay % ZN Actual zinc assay % PB Actual lead assay % CP CU/0.30 Actual % of chalcopyrite GNE Estimated gangue % (assumed correct) CU Actual copper assay % ZN Actual zinc assay % PB Actual lead assay %

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 129 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status CP CU/0.30 Actual % of chalcopyrite SP ZN/0.55 Actual % of sphalerite GA PB/0.90 Actual % of galena PYE Estimated pyrite % POE Estimated pyrrhotite % MTE Estimated magnetite %

TS 100-GNE Total sulphides

FE TS-(CP+SP+PB) Actual % of pyrite + pyrrhotite + magnetite

FEE PYE+POE+MTE Estimated % of pyrite+pyrrhotite+magnetite

SG = (GNE*2.7+(FE/FEE)*(PYE*5.0+POE*4.6+ MTE*5.18)+CP*4.1+ SP*4.0+GA*7.5)/100

The following SGs were assumed.

Assumed Specific Gravities

Mineral SG % of Metal Chalcopyrite 4.10 30 Sphalerite 4.00 55 Galena 7.50 90 Pyrite 5.00 Pyrrhotite 4.60 Magnetite 5.18 Gangue 2.70 Wardrop found there were good correlation between the physical SG measurements and the calculated SGs for the samples where both values could be determined. However the calculated SG is obviously heavily dependent on visual estimates by the geologists logging the drill core and consistency and accuracy of these estimates is always in question.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 130 of 132 Mineral Resources 2008 Explanatory Notes

Criteria Status

Classification • The classification of an Indicated Mineral Resource is as follows:

• Central Deposit - Search ellipse parameters of 58 x 52 x 35 metres - Minimum of three composites, and a maximum of two from any drill hole - Distance to closest composite is less than or equal to 30 metres • Inukshuk Deposit - Search ellipse parameters of 125 x 62.5 x 25 metres - Minimum of three composites, and a maximum of two from any drill hole - Distance to closest composite is less than or equal to 30 metres • North Deposit - Search ellipse parameters of 100 x 50 x 20 metres - Minimum of three composites, and a maximum of two from any drill hole - Distance to closest composite is less than or equal to 30 metres • Northwest Deposit - Search ellipse parameters of 145 x 81 x 20 metres - Minimum of three composites, and a maximum of two from any drill hole - Distance to closest composite is less than or equal to 30 metres • The remainder of the blocks within the interpreted three-dimensional wireframe was classified as being Inferred Mineral Resources.

Audits or • A desktop review of the Canadian Mineral Resources including Izok Lake Reviews was conducted by Behre Dolbear Australia (BDA) for Zinifex in 2008 and found the Mineral Resource estimate to be generally reasonable and appropriate. • The estimate calculated by Wardrop correlates well with previous estimates calculated by Strathcona Mining Services for Metall (1994). Mineral Resource Estimate Comparison Method Tonnes Cu % Zn % Pb % Ag g/t (,000) Wardrop ID &OK 14,770 2.54 12.77 1.25 71 Strathcona ID 14,738 2.50 13.00 1.30 68

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 131 of 132 Mineral Resources 2008 Explanatory Notes

11.5 Competent Person Statement

Tim Maunula is a member of the Association of Professional Geoscientists of Ontario and is a “Qualified Person” as defined in National Instrument 43-101 (NI 43-101) Standards of Disclosure for Mineral Projects of the Canadian Securities Administrators.

OZ Minerals Limited Mineral Resources Explanatory Notes - 30 June 2008 - FINAL.doc Page 132 of 132