April 8, 2005
[LOGO]
ROSCOE POSTLE ASSOCIATES INC.
Toronto, Ontario.
Vancouver, B.C.

		PAGE
SUMMARY		1
	Introduction and Conclusions		1
	Property Location, Access and Infrastructure		2
	History		2
	Geology and Mineralization		3
	Drilling and Sampling		3
	Mineral Resource and Mineral Reserve Estimates		4
		Mineral Resources		4
		Mineral Reserves		5
	Other relevant data and information		6
INTRODUCTION AND TERMS OF REFERENCE		8
DISCLAIMER		9
PROPERTY DESCRIPTION AND LOCATION		10
ACCESSIBILITY, CLIMATE, LOCAL RESOURCES, INFRASTRUCTURE AND PHYSIOGRAPHY		13
	Access		13
	Climate		13
	Infrastructure		13
	Physiography		13
HISTORY		15
	Exploration History		15
	Mining History		16
GEOLOGICAL SETTING		17
	Regional Setting		17
	Local Geology		19
	Property Geology		21
DEPOSIT TYPES		23
MINERALIZATION		23
EXPLORATION		23
DRILLING		24
	Small Diameter Core Drilling		27
	Large Diameter Core Drilling		27
	Large Diameter Drilling 2004		27
SAMPLING METHOD AND APPROACH		27
	Large Diameter Core Drilling		27
	Underground Bulk Sampling		28
	Diamond Valuations		28
SAMPLE PREPARATION, ANALYSES AND SECURITY		29
DATA VERIFICATION		29
ADJACENT PROPERTIES		30
MINERAL PROCESSING AND METALLURGICAL TESTING		30
MINERAL RESOURCE AND MINERAL RESERVE ESTIMATES		31
	General Statement		31
	Mineral Resource Estimate		31
		Database		31
		General Approach		32
		3D Volume Modeling		33
		Compositing		33
		Statistics		33
		Variography		33

		PAGE
Table 1 Diavik Mining Leases		10
Table 2 Diavik Diamond Mine Production		16
Table 3 Summary of Diavik Drilling		24
Table 4 Aber Adjacent Property Holdings		30
Table 5 Diavik Mineral Reserves December 31, 2004		31
Table 6 Diavik Additional Mineral Resources December 31, 2004		31
Table 7 Diavik Open Pit and Underground Mineral Reserves December 31, 2004		38
Table 8 Major Mining Equipment		39
Table 9 Mining Sequence		39
Table 10 Production Schedule		40
Table 11 Forecast Capital and Operating Costs		44
Table 12 Diavik Mine — Cash Flow Model		45
Table 13 Sensitivity Analysis (C$ Millions)		45

		PAGE
Figure 1 Diavik Mine Location Map		11
Figure 2 Diavik Mining Leases		12
Figure 3 Diavik Mine Site Plan		14
Figure 4 Regional Geology of Slave Structural Province		18
Figure 5 Local Geology of Lac de Gras Area		20
Figure 6 Property Geology — East Island Area		22
Figure 7 A154N Pipe Large Diameter Hole Locations		25
Figure 8 A154N Pipe Vertical Section A-A'		26

Roscoe Postle Associates Inc. (RPA) has been retained by Aber Diamond Corporation (Aber) to carry out an independent audit of the mineral reserves and mineral resources at the Diavik Diamond Mine (Diavik or the Diavik Mine) located in the Northwest Territories (NWT) of northern Canada. Aber Diamond Limited Partnership (40%) and Diavik Diamond Mines Inc. (60%) are joint venture participants in the Diavik Diamond Mine located at Lac de Gras in the Northwest Territories, Canada. Aber Diamond Limited Partnership is wholly owned, directly and indirectly, by Aber Diamond Corporation of Toronto, Canada. Diavik Diamond Mines Inc. (DDMI) is a wholly owned subsidiary of Rio Tinto plc of London, England and is the mine operator.

The Diavik Diamond Mine is adjacent to the Ekati Diamond Mine of BHP-Billiton which has been producing since 1998. The main operations at Ekati are located some 30 km northwest of the Diavik Diamond Mine site.

The Diavik Diamond Mine consists of an open pit mine, processing plant and supporting infrastructure. Diamonds produced at the site are transported to Yellowknife where they are sorted and distributed to the joint venture participants. By 2008 the mine and processing plant is expected to treat 2.5 million tonnes of kimberlite per year and forecasts in the order of 9 to 10 million carats of rough diamonds per year. Mining, which commenced in late 2002, is by open pit, with a transition to underground mining forecast scheduled to begin in 2008. Two open pits are planned on three separate kimberlite pipes. Exploration and development of the underground mine is scheduled to begin in 2005 and will involve the three kimberlite pipes below the open pit limits.

Mineral reserve and mineral resource estimates as of December 31, 2004 are tabulated below.

resource estimate. The mining, processing and economic parameters are reasonable and acceptable for conversion of mineral resources to mineral reserves. In RPA's opinion, the Diavik Diamond Mine mineral resources and mineral reserves dated December 31, 2004 are reasonable and acceptable.

RPA agrees with the classification of the DDMI mineral reserves into proven and probable categories. RPA has made one change to the classification of mineral resources: RPA has classified the A21 resources into indicated and inferred categories, whereas DDMI had all in the inferred category.

The Diavik Diamond Mine is located in the Northwest Territories some 300 km northeast of Yellowknife. Diavik Diamond Mines Inc. has five mining leases from the Government of Canada with a total area of 2,470.5 hectares. Theses leases are granted under the terms specified in the Territorial Lands Act and the Territorial Lands Regulations.

A 1,600-metre long airstrip affords regular access to the Diavik Diamond Mine by fixed wing aircraft. Personnel are transported to and from the site, from a number of northern communities, by small commuter aircraft. Weekly service to and from Edmonton is provided by way of Boeing 737 aircraft. The Diavik Diamond Mine is accessible in winter by truck, via a 425 km long winter road, typically in operation from January to March. The majority of supplies required for the mine, including, fuels, lubricants and explosives, are transported over this road.

To support the mining and processing operation at the Diavik Diamond Mine, infrastructure at the site includes a permanent accommodation 264-room complex, part of a 700-person construction camp for overflow accommodation, maintenance shops, warehouse, administration offices, geochemical and environmental laboratory, three 700 hp diesel powered boilers for heating, and a power generating facility consisting of five 4.4 MW diesel generators. Waste heat is recovered from the generators for general heating requirements.

The Lac de Gras region is north of the tree line in the barren lands and is characterized by an abundance of small to large, shallow lakes, impeded drainage, low relief, and mix of hummocky boulder strewn terrain and rock exposures. In recent history, Dene, Metis and Inuit peoples used the Lac de Gras area for subsistence hunting and trapping. Today, human activity in the Lac de Gras area is largely confined to exploration, mining, outfitting and guiding related activities.

The Diavik Diamond Mine, known as the Diavik Diamond Project prior to the commencement of commercial production, was created to explore and develop diamondiferous kimberlite pipes in an area approximately 300 km northeast of Yellowknife. The original Diavik claims were staked by Aber in late 1991 and early 1992. In 1992 a joint venture was formed between Aber Resources Ltd. (now Aber Diamond Corporation) and partners and Kennecott Canada Inc. (now Kennecott Canada Exploration, Inc.) to continue exploration of the Diavik claims.

Early exploration relied on airborne geophysical surveys and heavy mineral sampling in till. Prospective targets were prioritized for follow up geophysics, sampling and drilling to test for kimberlite, delineate the kimberlite bodies and determine their micro-diamond contents. If results were encouraging, large diameter core holes were drilled to obtain mini-bulk samples.

Four potentially economic kimberlite pipes were discovered by the Diavik joint venture under the waters of Lac de Gras adjacent to East Island. Mini-bulk samples were obtained from the A154S, A154N, A418 and A21 pipes by large diameter core drilling. Underground bulk sampling of the A154S and A418 pipes was undertaken via a decline driven from the shore of East Island. The mini-bulk samples and the underground bulk samples were processed at a pilot plant to recover diamonds for mineral resource estimation and valuation of representative parcels from each pipe.

Diavik Diamond Mines Inc. (DDMI), which was assigned the Kennecott interest, was established in 1996 to develop the joint venture prospects. Based on a definitive Diavik Feasibility Study by SNC Lavalin in 2000, a

production decision was taken to develop the Diavik Diamond Mine. Construction proceeded through 2001 to 2002. Equipment, construction materials and fuel were trucked to the site on the winter road.

A kimberlite processing and diamond recovery plant was constructed along with associated infrastructure noted above, plus fuel storage tanks, processed kimberlite containment area, and water storage and treatment facilities. A 3.8 km long water retention dike was constructed around the planned site of the A154 open pit. After dredging of the lake bottom sediments and dewatering the diked area, till overburden was removed to expose the A154S and A154N pipes for mining.

Mining and processing of kimberlite commenced in late 2002. Kimberlite mined and processed has been mostly from the A154S kimberlite with some contribution from A154N. Production to the end of 2004 totals 11.4 million carats of diamonds from 3.1 million tonnes of processed kimberlite for a recovered grade of 3.7 ct/t.

The Diavik Diamond Mine area is located near Lac de Gras in the central part of the Slave Structural Province which forms a distinct cratonic block within the Canadian Precambrian Shield. The Slave craton contains deformed and metamorphosed, Archean aged metaturbidite and lesser metavolcanic rocks of the Yellowknife Supergroup. These supracrustal rocks have been intruded by extensive Archean granitoids, and are in turn intruded by undeformed, late Archean granites and diabase dike swarms. Pleistocene continental glaciation and retreat has left a thin, discontinuous mantle of sandy and bouldery basal and ablation tills and ice contact deposits such as eskers in the area.

Local geology in the Lac de Gras area consists of three Archean lithologies: (1) greywacke-mudstone metaturbidites, (2) biotite±hornblende tonalite to quartz diorite, and (3) two-mica or K-spar porphyritic granite and granodiorite. The four Diavik kimberlite pipes are aligned in a northeast-southwest direction along with other pipes. Country rocks to the kimberlites are muscovite-biotite granites cut by pegmatite, locally with inclusions of biotite schist (metamorphosed turbidites).

The Diavik kimberlites were formed by volcanic surface eruptions and near surface injections of kimberlite magma and volcaniclastic debris into the granitic country rocks and into mid Cretaceous to Tertiary mudstones that once covered the Archean basement. The kimberlites contain a number of facies, broadly classed into three types of material. Hypabyssal facies represents material crystallized from kimberlite magma. It generally represents a small portion of the Diavik pipes along walls and in roots. Volcaniclastic material formed by fallback or slumping into a crater and epiclastic material composed of kimberlitic and mudstones washed into the crater. The volcaniclastic and epiclastic material make up the bulk of the Diavik kimberlites. Exotic fragments consisting of Cretaceous to Tertiary mudstones and granitic country rocks occur in kimberlite. Mudstone in places forms a significant portion of the kimberlite and dilutes the diamond grade.

The four Diavik kimberlite pipes for which mineral resources and mineral reserves are reported have been delineated and sampled by a combination of small diameter core drill holes, large diameter core drill holes, large diameter reverse circulation drill holes and Sonic drill holes. Small diameter core holes (HQ and NQ size) drilled from 1994 to 1998 totals 19,494 m in 71 holes on the four pipes. Large diameter core holes (152 mm and 85 mm diameter) from 1996 to 1998 total 11,746 m in 38 holes on the four pipes. An additional 27 small diameter core holes were drilled on the A154S and A154N pipes in 2003 and 2004 (4,951 m) along with ten large diameter reverse circulation holes (61 cm diameter) on A154N and nine 35 cm large diameter reverse circulation and 152 mm Sonic holes on the A418 pipe.

Mini-bulk samples from the large diameter core holes were processed at a pilot plant. Overall for the four Diavik pipes, a total of 266 t were processed from the four pipes to recover 1,028 ct of diamonds. Underground bulk samples were processed at two pilot plants. The 2,587 t sample from A154S yielded 12,688 ct and the 3,350 t sample from A418 yielded 8,325 ct. Similarly, mini-bulk samples from large diameter reverse circulation holes drilled in the A154N pipe in 2004 were processed at a pilot plant. A total of 853 t were processed to recovery 2,109 carats from the A154N pipe.

Diamond valuations for the four pipes were carried out for the 2000 and prior Feasibility Studies. For a representative parcel of the A154S bulk sample diamonds, the value was US$79 per carat. For representative samples of the A418 bulk sample diamonds, different values were obtained for different kimberlite types: US$69 per carat for Type A and US$49 per carat for Type B. Relatively small parcels of diamonds from the large diameter holes returned values of US$33 per carat for A154N and US$28 per carat for A21. The A154N value is superseded by a valuation of US$82 per carat on 11,771 carats recovered from a bulk sample mined in 2003.

In RPA's opinion, the drill hole and bulk sampling data are valid and appropriate for estimation of mineral resources and mineral reserves.

There are some significant changes from previous reserve and resource statements. The A154N mineral reserve has increased significantly: previous mineral resources below the open pit has been upgraded by more drilling and have been demonstrated to be economic based on a higher diamond value obtained in bulk sampling. The A21 pipe mineral reserves have been downgraded to mineral resources because of uncertainty in the economics in the 2004 estimate.

The general approach taken by DDMI is simple kriging of stone density values using results of the large diameter core drilling and large diameter reverse circulation drilling. Simple kriging weights grade by domain means as well as by the spatial distribution of composites, in contrast to ordinary kriging or other interpolation methods that weight grades only by spatial distribution. The stone density values for each block were multiplied by the average stone weight (ct/stn) for the particular kimberlite unit (domain) to obtain grade values (ct/t). The block volumes were multiplied by the average bulk density for the particular kimberlite domain to obtain block tonnage.

The large diameter core samples were composited to regularize the samples for uniform weighting prior to geostatistical interpolation of stones/tonne values. The composites were declustered in preparation for block model interpolation. Block models were constructed for each pipe with block dimensions of 15 m by 15 m horizontal by 10 m vertical. The blocks corresponded vertically to the planned mining bench heights.

RPA audited the Diavik diamond mineral resource estimates in connection with a previous assignment in 2001. At that time, RPA's opinion was that the DDMI mineral resource estimate was reasonable and acceptable. Since that time, the Diavik Diamond Mine has come into operation and has produced significant quantities of diamonds. Production has come primarily from the A154S pipe with some from the A154N pipe.

Reconciliation studies by DDMI compared tonnes and grade mined with reserve estimates. Several factors were identified as contributing to lower than expected production grades in 2003, including presence of mud rich kimberlite, higher moisture content and lower density than expected in the upper benches of A154S, lower than expected recovery of small diamonds in the process plant, and processing of some low grade A154N kimberlite. These factors were taken into account in the December 31, 2004 resource and reserve estimate and led to the introduction of a 94% metallurgical recovery factor for the processing plant.

For the December 31, 2004 resource estimate, DDMI has redone the 2000 block model using the same simple kriging methodology as described above, but with results of more drilling, bulk sampling, operating experience and the factors noted above.

RPA reviewed the 2004 DDMI mineral resource estimate and did an independent block model estimate of the A154N kimberlite pipe, the pipe with the most new data collected in 2003 and 2004. RPA used ordinary kriging of ct/t values in LDC drill holes instead of simple kriging of stn/t values used by DDMI. Results of the RPA estimate compared closely with the 2004 DDMI mineral resource estimate for the A154N pipe.

In RPA's view, the DDMI mineral resource estimates for the A154S, A154N, A418 and A21 pipes are reasonable and acceptable.

DDMI classified mineral resources in each kimberlite pipe into measured, indicated and inferred categories depending on the density of drilling information, sample spacing, and assessment of geological continuity. DDMI classifies its mineral resources and mineral reserves under the Australian Joint Ore Reserve Committee (JORC) code. RPA has reviewed the classification of the Diavik mineral resources with respect to the CIM standards for classification of mineral resources, as required under National Instrument 43-101 (NI 43-101).

RPA agrees with the DDMI mineral resource classifications with the exception of the A21 pipe, which DDMI classified all as inferred. In RPA's view, it is more appropriate to classify the A21 mineral resources as indicated and inferred under the CIM standards of classification, to reflect level of geological knowledge and confidence in the grade and tonnage estimates. In the 2000 Feasibility Study, the A21 Pipe was classified as indicated resource and inferred resource; most of the indicated resource was converted to probable reserve.

Pre-production mineral reserves were estimated in the 2000 Feasibility Study from the mineral resources and included both open pit and underground mineral reserves. Since commencement of diamond production in late 2002, the original mineral resource estimate, the open pit designs and the underground mining plans have been modified from those in the Feasibility Study. The modifications to the A154S, A154N and A418 open pit designs are relatively minor. The A21 pipe was originally in the mine plan as an open pit and classified as mineral reserves, but is not included in the 2004 mineral reserves because of uncertain economics. The A21 pipe has been declassified to mineral resources, as noted previously. DDMI plans to do more work in future on the A21 pipe to gain more confidence in the mineral resource estimate and diamond value.

New studies have been undertaken on underground mining of the A154S, A154N and A418 pipes which have resulted in major changes to the underground mineral reserves. The largest change is to the A154N pipe which had no underground mineral reserves in the 2000 Feasibility Study, but has a significant underground mineral reserve in 2005. This change is largely due to a new valuation of A154N diamonds based on a bulk sample from the open pit in 2003 and samples from large diameter reverse circulation drilling in 2004.

External dilution is estimated at 4%, primarily to allow for inclusion of granitic wallrock in the kimberlite mill feed. Reconciliation studies indicate that 4% is a reasonable estimate. There is no internal dilution since all of the kimberlite will be mined; similarly mining recovery is assumed to be 100%. The concept of a cut-off grade is not applicable since there will be no mining selectivity: all of the kimberlite will be mined to the economic depth of each open pit.

DDMI revised and reassessed the 2000 Feasibility Study underground mine designs and parameters in 2004. Modifications were made for the A154S and A418 pipes where underground mining is planned to take place after most of the open pit mining is completed. In the case of the A154N pipe, where underground mining was not planned in the 2000 Feasibility Study, underground mining is now planned because a new diamond valuation makes the economics more attractive.

Underground mining is planned to be by underhand cut and fill, cut and fill, and blast hole cut and fill with access via haulage ramps with sublevel development. Major considerations in mine design are the inherent weakness of the kimberlite and water inflow. Mining recovery and dilution estimates are included in the reserve estimate. Dilution is estimated to be 4% at zero grade for the open pits and 11% to 16% at zero grade for underground mining, depending on the mining method applied. Mining recovery is assumed to be 100% in the open pits and 90% for underground mining.

In 2005, DDMI is commencing a program of underground exploration and test mining at the A154S, A154N and A418 pipes to test ground and water conditions in order to do more detailed mine planning. In RPA's view, this program is justified and should result in much valuable information.

In its determination of mineral reserves, DDMI assessed the economics of open pit mining and underground mining. With over two years of operating experience, parameters such as direct and indirect operating costs, ongoing capital requirements, mining productivity, processing plant capacity and recovery, and

diamond values. Other parameters, such as underground mining capital and operating costs, dilution and recovery, productivity, ground conditions, water inflow, have been reviewed and reassessed since the time of the 2000 Feasibility Study. RPA has reviewed the economic studies and parameters used by DDMI to justify the designation of mineral reserves and concurs that they are reasonable and acceptable.

RPA concurs with the mineral reserve estimate and the classification into proven and probable categories, as listed in the following table.

Currently, mining of the Diavik Diamond Mine is by conventional open pit methods in the A154 open pit. Ore mined from the A154S and A154N pipes in ten metre benches is loaded into trucks with hydraulic shovels and hauled to the plant and stockpiled for rehandling into the plant. Ore is fed to the processing plant in batches. The mine trucks are used to haul waste to the waste dumps. The mine operates around the clock; most operators work a 12-hour shift.

Production from the A418 open pit is scheduled to commence in 2008. Construction on the A418 water retention dike is scheduled to start in 2005. The 2005 Mine Plan includes only Proven and Probable Mineral Reserves.

Underground mining of parts of the A154S, A154N and the A418 pipes below the completed open pits is scheduled to begin in 2008. An underground exploration and test mining program will begin in 2005 to further delineate and test underground stoping methods. Stoping methods planned include underhand cut and fill, cut and fill, and blast hole cut and fill.

The processing plant recovers diamonds from the kimberlite by means of primary crushing and scrubbing to remove fine material to tailings, secondary crushing to limit the particle top size to 25 mm, dense medium separation (DMS) to recover heavy minerals including diamonds, re-crushing DMS reject to 6 mm to liberate locked diamonds, wet and dry x-ray sorting plus magnetic separation to recover diamonds from DMS concentrate, single particle x-ray sorting, sizing, packaging and transport to Yellowknife. Processed kimberlite is placed in the processed kimberlite containment storage area.

Diamonds are flown from the Diavik Diamond Mine to the production splitting facility (PSF) in Yellowknife where the diamonds are cleaned, sorted and split into Aber's 40% share and DDMI's 60% share. Aber's share of the diamonds is transported to Toronto for further sorting and then sale into international markets.

DDMI completed a thorough environmental assessment before the Diavik Diamond Mine site was developed. An environmental staff of nine is responsible for monitoring, directing, reporting and communicating on the environmental issues. DDMI reports that the project is in compliance with all permits and that there are no other environmental liabilities known at this time.

Capital costs are estimated at some C$869 million from 2005 to 2017. A large part of the capital expenditure is scheduled over the next four years and relates to the underground program and construction of the A418 dike. Annual operating costs are in the order of C$220 million currently.

In order to demonstrate the economic viability of the Diavik mineral reserves, RPA has prepared a cash flow model for the Diavik Diamond Mine and done sensitivity analysis. RPA has used DDMI forecasts of mine production, capital costs and operating costs from 2005 to 2017. Diamond values are from the 2000 Feasibility Study except for the A154N pipe for which the 2003 bulk sample value is used. For the purposes of the model, the Diavik Diamond Mine was assumed to be a taxable entity. This is not actually the case, since each of the Diavik joint venture participants is responsible for its own taxes.

The RPA cash flow model is very robust and clearly shows that the Diavik Diamond Mine is profitable and that the designation as mineral reserves is justified.

Roscoe Postle Associates Inc. (RPA) has been retained by Aber Diamond Corporation (Aber) to carry out an independent audit of the mineral reserves and mineral resources at the Diavik Diamond Mine (Diavik Mine) located in the Northwest Territories (NWT) of northern Canada. Aber Diamond Limited Partnership (40%) and Diavik Diamond Mines Inc. (60%) are joint venture participants in the Diavik Diamond Mine located at Lac de Gras in the Northwest Territories, Canada. Aber Diamond Limited Partnership is wholly owned, directly and indirectly, by Aber Diamond Corporation of Toronto, Canada. Diavik Diamond Mines Inc. (DDMI) is a wholly owned subsidiary of Rio Tinto plc of London, England and is the mine operator.

RPA is familiar with the Diavik Diamond Mine through its role as Independent Engineer for a group of Canadian and international banks that financed Aber's share of the capital cost of developing the mine. RPA and Hatch Associates Ltd. (Hatch) were retained by a bank group in 2001 to carry out an initial technical audit and ongoing monitoring of the Diavik Diamond Project when it was an early stage of construction. RPA/Hatch prepared an initial due diligence report, including an audit of mineral resources and mineral reserves, and monitored the project through the construction period into production which started in late 2002. In early 2003, RPA/Hatch carried out another due diligence review in conjunction with a refinancing of the Aber loan.

Aber reported mineral reserves and resources at the Diavik Diamond Mine as of December 31, 2004 in a press release dated March 9, 2005. A key document that supports the mineral reserves is the Diavik Joint Venture 2005 Workplan Budget and 2006 - 2009 Forecast (2005 Mine Plan) dated October 2004. This document outlines the mine and plant production, operating and capital cost for the operation from 2005 to 2009.

The Diavik Diamond Mine consists of an open pit mine, processing plant and supporting infrastructure. Diamonds produced at the site are transported to Yellowknife where they are sorted and distributed to the joint venture participants. By 2008 the mine and processing plant will treat 2.5 million tonnes of kimberlite per year to produce in the order of 9 to 10 million carats of rough diamonds per year. Mining, which commenced in late 2002, is by open pit, with a transition to underground mining forecast to begin in 2008. Two open pits are planned on three separate kimberlite pipes. Exploration and development of the underground mine is scheduled to begin in 2005.

All of the infrastructure, including the airstrip, and the process plant are located on East Island of Lac de Gras. Access for personnel and light cargo is by air. Fuel and heavy equipment are trucked to the site over a winter road which operates for 8 to 10 weeks every year. The capital budget for construction of the Diavik Diamond Mine, processing plant and supporting infrastructure was approximately C$1.3 billion.

One unique feature of this mine is that all of the kimberlites mined are located under the waters of Lac de Gras near East Island. A water retention dike was built extending from East Island around two kimberlite pipes and water remaining inside the dike was pumped out prior to mining.

RPA reviewed data and held discussions with appropriate personnel at the Aber office in Toronto, the Diavik office in Yellowknife and the Diavik Diamond Mine.

RPA personnel involved in this audit, along with their areas of specialization, are:

All three RPA consultants visited the Diavik site on November 23 and 24, 2004 and the DDMI Product Splitting Facility (PSF) in Yellowknife on November 25, 2004. Data were collected and discussions were held with the following DDMI personnel:

All of the personnel were very co-operative, helpful and professional, and provided RPA with information requested.

Metric units are used throughout this report. All dollar amounts are Canadian currency (C$) unless otherwise noted; for example, diamond values are in US$. Abbreviations used in this report are as follows:

M		million
m		metres
m3		cubic metres
mm		millimetres
km		kilometres
ct		carats
t		tonnes
stn		stone or stones
stn/t		stones per tonne
ct/t		carats per tonne
ct/stn		carats per stone
mRL		elevation above mean sea level in metres
tpd		tonnes per day
ppm		parts per million

This report has been prepared by Roscoe Postle Associates Inc. (RPA) for Aber. The information, conclusions, opinions, and estimates contained herein are based on:

RPA does not guarantee the accuracy of conclusions, opinions, or estimates that rely on third party sources for information that is outside the area of technical expertise of RPA. For property title, environmental and permitting aspects, and contracts entered into by DDMI, RPA has relied on information supplied by DDMI and Aber.

The Diavik Diamond Mine is located in the Northwest Territories some 300 km northeast of Yellowknife (Figure 1). It is at latitude 64^o29'50" N and longitude 110^o 20' 21"W.

DDMI has five leases from the Government of Canada to mine the area shown in Figure 2. Theses leases are granted under the terms specified in the Territorial Lands Act and the Territorial Lands Regulations. The lease numbers, areas and annual payments are shown in the Table 1. These leases carry certain obligations for reclamation and other requirements, and require certain security deposits.

Lease #	Lease Date	Expire date	Yearly payments
76D/8-5-2	March 29, 2000	March 28, 2030	$	8,493.00	339.7
76D/8-6	March 29, 2000	March 28, 2030	$	18,812.00	752.5
76D/8-7	March 29, 2000	March 28, 2030	$	21,180.00	803.4
76D/9-5-2	March 29, 2000	March 28, 2030	$	6,915.00	276.6
76D/9-9	March 29, 2000	March 28, 2030	$	7,458.00	298.3
Total					2,470.5

A 1,600-metre long airstrip affords regular access to the Diavik Diamond Mine by fixed wing aircraft. The airstrip is capable of accommodating large transport and passenger aircraft. Personnel are transported to and from the site, from a number of northern communities, by small commuter aircraft. Weekly service to and from Edmonton is provided by way of Boeing 737 aircraft. The Diavik Diamond Mine is a remote site with strictly controlled access. All personnel working at the site, contractors and visitors travel to the site on regularly scheduled flights and charter flights from Yellowknife, Edmonton, and other northern communities.

The Diavik Diamond Mine is accessible, in winter, by truck, via a 425 km long "ice road." This winter road, typically in operation from January to March, links a number of mining, exploration and outfitting camps with Yellowknife. The majority of supplies required for the mine, including, fuels, lubricants and explosives, are transported over this road.

The mean annual temperature at the Diavik Property is –12 degrees Celsius. Temperatures in the summer months may, however, be in excess of 25 degrees Celsius. The average minimum temperature is –35 degrees Celsius in January, although extremes may be below –50 degrees Celsius. The mean annual precipitation is approximately 375 mm, 60% of which falls as snow. The project site is generally windy with the 100 year extreme exceeding 90 km / hr. Available daylight ranges from a minimum of four hours per day in December to a maximum of 22 hours in June.

To support the mining and processing operation at the Diavik Diamond Mine the following infrastructure has been constructed at the site (Figure 3):

The Lac de Gras region is north of the tree line in the barren lands and is characterized by an abundance of small to large, shallow lakes, impeded drainage, low relief, and mix of hummocky boulder strewn terrain and rock exposures. Lac de Gras itself varies from 4 m to 25 m deep in the area of the Diavik pipes, and forms the headwaters of the Coppermine River system.

From 1994 to 1997, extensive baseline studies of the local and regional environment were carried out to develop an environmental knowledge base for the Diavik Diamond Mine.

Archaeological studies have found evidence of past human use of the area. In the Lac de Gras region, archaeologists and their assistants have found quarries, artifact scatters and isolated artifact finds. On the east island, primarily quarried veins of quartz and numerous quartz chip scatters were found. In more recent history, Dene, Metis and Inuit peoples used the Lac de Gras area for subsistence hunting and trapping. Today, human activity in the Lac de Gras area is largely confined to exploration, mining, outfitting and guiding related activities. Communities that would benefit by the project include Wekweti, Gameti, Wha Ti, Rae-Edzo, Dettah, N'dilo and Lutsel K'e, as well as the Metis of the North Slave.

Typical of arctic lakes, aquatic productivity in Lac de Gras is low. This is the natural result of relatively low concentrations of nutrients, low light levels during the winter months, extended periods of ice cover and low water temperatures. Lake trout, cisco, round whitefish, arctic grayling, burbot, longnose sucker and slimy sculpin are among the fish species found in Lac de Gras. The Lac de Gras area is a mosaic of landscape features and habitat types that support many wildlife species. Eighty-four bird species and 16 mammal species have been confirmed as summer visitors or permanent residents in the area. Half of the bird species regularly breed in the area, with the remainder being migrants or uncommon visitors. Lac de Gras falls within the extensive range of the Bathurst caribou herd, part of which migrates through the area during spring and fall. Up to 100,000 caribou have been observed in the Lac de Gras region during spring migration and, in some years, a large number pass close to the project site. Wolves, following these caribou, also den in the area during summer. The home ranges of an estimated 30 adult and sub-adult barren-ground grizzly bears overlap in the Lac de Gras region.

While the region sees many visitors, on the East Island only a few species are permanent residents. These include red fox, arctic hare, arctic ground squirrels, red-backed voles, brown lemmings and rock ptarmigan. In some years, small to large numbers of caribou may cross the ice of Lac de Gras to the East Island during spring migration and may visit the islands for short periods during summer and fall movements. Grizzly bears, wolves and wolverines have large home ranges and occasionally visit the East Island, especially when following caribou herds. There are no recent grizzly bear dens or permanent wolf denning sites on the island. Many different species of birds have been observed to stop briefly at the East Island during spring and fall migrations.

The gently rolling tundra surrounding the Diavik Diamond Mine, commonly called the barren lands because of the lack of trees, is made up of numerous lakes, bedrock outcrops, and glacially distributed boulder fields, till, eskers. There is very little soil in the area and is a continuous permafrost zone. Vegetation cover in the region is typical of the barren lands and includes dwarf birch, northern Labrador tea, blueberry and mountain cranberry species, with willow, sphagnum moss and sedge tussocks dominating the low-lying wet areas.

The Diavik Diamond Project was created to explore and develop diamondiferous kimberlite bodies (pipes) located between latitudes 64^o 29' and 64^o 31' N, and between longitudes 110^o 0' and 110^o 13' W, approximately 300 km northeast of Yellowknife in the Northwest Territories. The original Diavik claims were staked by Aber in late 1991 and early 1992. In 1992 a joint venture was formed between Aber and partners and Kennecott Canada Inc. (now Kennecott Canada Exploration, Inc.) to continue exploration of the Diavik claims.

Exploration relied on airborne geophysical techniques and heavy mineral sampling in till. These methods identified prospective targets, which were then prioritized for additional exploration by more detailed geophysics and sampling. The most prospective targets were then drilled to define the extent of the kimberlite (delineation drilling) and for micro-diamond determination. If results were encouraging, large diameter core holes were drilled to obtain mini-bulk samples (6 in. diameter core to 250 m followed by PQ (3.5 in. diameter) to the end of hole). Four potentially economic kimberlite pipes were discovered by the Diavik joint venture under the waters of Lac de Gras adjacent to East Island.

Pipe A21, the first of the four pipes to be discovered, was first drilled in April 1994. Pipe A154S was discovered in May 1994 and pipe A154N was first drilled in the fall of 1994. Pipe A418 was discovered in the spring of 1995. All four kimberlite pipes display distinct electromagnetic signatures. A154N and A21 also have coincident weak magnetic lows. Mini-bulk sampling with large diameter core (LDC) holes began in 1995.

Underground bulk sampling of the A154S and A418 pipes was undertaken via a decline driven from the shore of East Island. The ramp was collared in permafrost but proceeded into unfrozen ground under the lake and in parts of the kimberlite pipes. Ground support using rock bolting, screening and shotcreting was required to control water inflow and erosion of the kimberlite, especially in unfrozen ground.

Mini-bulk samples from the LDC drill holes and the bulk samples were processed at the DDMI pilot plant in Yellowknife. Approximately half of the A154S bulk sample was processed in the Koala plant at the Ekati diamond mine, located 35 km northwest of Diavik. A total of 266 t of LDC material was processed from the four pipes and returned 1,028 ct of diamonds. The bulk sample of 5,937 t from the A154S pipe returned 12,688 ct of diamonds and the bulk sample of 2,587 t from the A418 pipe returned 8,325 ct of diamonds, with a bottom screen size of 0.85 mm. Valuations of the diamonds were carried out on representative samples of the packages.

DDMI, which was assigned the Kennecott interest, was established in 1996 to develop the joint venture prospects. As noted previously, DDMI, a wholly owned subsidiary of Rio Tinto plc, owns 60% of the Diavik Diamond Mine and is the operator and Aber owns 40%. Four kimberlite pipes were considered for development: A154S, A154N, A418, and A21.

From September 1996 to June 1998, DDMI completed a series of resource estimates on the A154S, A154N, A418 and A21 pipes. In 1999 and 2000 the DDMI mineral resource and reserve estimates were audited and estimated independently by Mineral Resources Development, Inc. (MRDI). The June 1998 DDMI resource estimate by was used as the basis for Diavik feasibility studies, including the definitive Diavik Feasibility Studies by SNC Lavalin in 2000 (2000 Feasibility Study). A production decision was taken to develop the Diavik Diamond Mine and construction started in 2001.

Construction of the Diavik Diamond Mine proceeded through 2001 to 2002. Equipment, construction materials and fuel were trucked to the site on the winter road which is established for about a ten week season every January to March from Yellowknife to the Lupin Gold Mine north of Diavik.

Site facilities include a kimberlite processing and diamond recovery plant, diesel power generation plant, maintenance shops, office complex, accommodation and recreation facilities, fuel storage tanks, processed kimberlite containment area, water storage and treatment facilities and an airstrip.

A 3.8 km long water retention dike was constructed around the planned site of the A154 open pit. After dredging of the lake bottom sediments and dewatering the diked area, till overburden were removed to expose the A154S and A154N pipes for mining.

Mining and processing of kimberlite commenced in late 2002. Kimberlite mined and processed has been mostly from the A154S kimberlite with some contribution from A154N. Mining of the A154S pipe started on the 385 m RL bench and was on the 320 m RL bench in late 2004. Mining of the A154N pipe started on the 370 m RL bench and was on the 360 m RL bench in late 2004. Table 2 lists Diavik production by year from start-up in November 2002 to the end of 2004.

Year	000s Tonnes Processed	000s Carats Recovered	Carats per Tonne
2003*	1,175	3,833	3.3
2004	1,950	7,575	3.9

Total	3,125	11,408	3.7

The Diavik Diamond Mine area is located near Lac de Gras in the central part of the Slave Structural Province which forms a distinct cratonic block within the Canadian Precambrian Shield. The Slave craton contains deformed and metamorphosed, Archean aged metaturbidite and lesser metavolcanic rocks of the Yellowknife Supergroup. These supracrustal rocks have been intruded by extensive Archean granitoids, and are in turn intruded by undeformed, late Archean granites and diabase dike swarms. Figure 3 depicts the regional geology of the Slave Structural Province.

Structurally the Slave craton is dominated by Proterozoic faulting and emplacement of several sets of diabase dikes (Stubley, 1998). These include the north-northwesterly trending Mackenzie dike swarms, east-northeast trending Malley and east-west trending Mackay dikes may have played a role, at least empirically, in the structural ground preparation that allowed for the emplacement and localization of the kimberlite pipes in the region.

More recently, the Wisconsin continental glaciation and retreat, about 8,000 to 14,000 years ago, has left a thin, discontinuous mantle of sandy and bouldery basal and ablation tills and ice contact deposits such as eskers in the area. Younger glaciolacustrine sediments are preserved locally in topographic depressions.

The Slave Geological Province may be divided into east and west domains based on isotope age data and deep crustal geophysical studies that suggest that the west domain represents older, thicker and more resistive lithosphere than the east domain that underlies the Lac de Gras area. Based on reported kimberlite exploration successes to date, the east Slave area has been more productive in that it hosts most of the over 200 pipes discovered to date in the NWT and Nunavut.

Most reported diamondiferous kimberlite pipes found in the Slave Province are relatively small compared to kimberlite occurrences world wide. They generally have a surface area of about two hectares (5 acres). The Diavik pipes have similar dimensions.

Three main Archean lithologies are present in the Lac de Gras area (Stubley, 1998). These are:

All of these rock types are present on East Island, as shown in Figure 5, local geology of the Lac de Gras area. Stubley (1998) notes that regional aeromagnetic patterns and limited lake bottom drill hole information suggest that metaturbidites underlie much of Lac de Gras to the south and west of East Island. Stubley (1998) interprets the tonalite suite to be a small stock centred on southern East Island.

Figure 6 shows the geology and structural features of the Diavik mining property. The four kimberlite pipes containing the Diavik mineral reserves and mineral resources are situated off the east and southeast shore of East Island in Lac de Gras (A154S, A154N, A418, and A21). The four pipes are aligned in a northeast-southwest direction along with other pipes. The A154N, A154S and A418 pipes are clustered together 3 km to 4 km northeast of A21.

The area of the cluster is underlain by late Archean (2.5 to 2.8 billion years) muscovite-biotite granites. The granites are cut by pegmatites locally, have inclusions of Yellowknife Group biotite schist and are variably jointed and fractured sub vertically and sub horizontally. The basement in the A21 area is similar. A kilometre wide belt of northwest trending Yellowknife Group greywacke, siltstones and mudstones occupies an area on East Island between A21 and the other pipes.

Proterozoic diabase dikes representing each of the five sets and ages (1.2-2.2 billion years) known in the region have been traced on the Diavik property by airborne and ground magnetic surveys, and intersected in drilling and underground development work. These dikes have blocky fracture and joint patterns that may provide permeable channel ways for groundwater.

A fault has been encountered in mining of the A154S and A154N pipes, named Dewey's Fault. It trends northeasterly, dips steeply and passes through both kimberlite pipes in the open pit. In places it contains kimberlite dike material.

The Diavik kimberlites are Eocene in age (54-58 million years) and were formed by volcanic surface eruptions and near surface injections of kimberlite magma and volcaniclastic debris into the granitic country rocks and into mid Cretaceous to Tertiary mudstones. These latter formations covered the Archean basement at the time of kimberlite emplacement but have since been entirely eroded from the area.

Kimberlites are heterogeneous alkalic ultramafic rocks exhibiting a wide variety of textures. They have variable matrix to clast ratios, variable matrix alteration (serpentinization, clay alteration), and often contain country rock inclusions, mantle xenoliths and xenocrysts of various unique mineral phases including diamond. Kimberlite within the pipes at Diavik is broadly classed into three types of material.

The deposits being mined at Diavik are diamond bearing kimberlite pipes, which form within thick Precambrian cratons. Since discovery of the Ekati diamondiferous kimberlites in 1991, some 200 kimberlite pipes have been found in the Slave craton in the Northwest Territories and Nunavut. Of these, several are considered worthy of mining at the Ekati Mine and four at the Diavik Diamond Mine in the Lac de Gras area of the NWT. The Jericho pipe in Nunavut is being developed by Tahera Corporation and the Snap Lake kimberlite dike is being developed by De Beers in the NWT.

In a simple model, diamonds are included as xenocrysts in a kimberlite magma as it is formed and ascends through the upper mantle and crust. Upon nearing the surface, the kimberlite magma, which is rich in volatiles such as CO₂, explosively erupts, forming the characteristic "carrot-shaped" diatreme. Abundant kimberlite is erupted as pyroclastic ejecta and falls both within and adjacent to the pipe. The diatreme is filled with a combination of pyroclastic kimberlite, hypabyssal kimberlite, and in the case of the Lac de Gras pipes, Cretaceous mudstone that slumped back into the pipe. At Lac de Gras, the tops of the pipes were removed by continental glaciation. The kimberlites, being softer than the surrounding rocks, tended to form depressions, which after the glaciers retreated, filled with water, forming small lakes. When the pipes occur under larger lakes, such as Lac de Gras, the pipes typically lie beneath small depressions on the lake bottom.

The diamondiferous kimberlite pipes at Diavik intrude granitic rocks and metaturbidites of the Slave craton. The Diavik pipes vary from 118 m to 158 m in diameter at surface and are circular to slightly elongated in the northeast-southwest direction with bulges occurring along the margins. Their surface areas are in the range of one to two hectares. In cross section the pipe walls are near vertical to steeply inward dipping and form typical kimberlite carrot shapes extending to and narrowing slightly at depth. A154S and A21 pipes are carrot shaped to "champagne glass" flute shaped and begin to narrow below 315 m depth below surface. A418 and A154N are nearly cylindrical to the depths drilled.

Deepest drilling completed on these pipes using large diameter holes reached the –116 m RL level for A154S, or 531 m below surface, –44 m RL for A154N (459 m depth), –37 m RL for A418 (452 m depth) and +44 m RL for A21 (371 m depth). The pipes have been modelled to –125 m RL for A154S to 125 m RL for A154N to –95 m RL for –A418, and –95 m for A21.

The kimberlite within each pipe has been subdivided into 4 to 7 units for resource modelling. Units were broadly defined with the purpose of correlation across the pipe on a mine scale. The units was defined on the basis of macroscopic criteria, mud dilution, grain size, magnetic susceptibility, and textural and alteration characteristics. These aspects of kimberlite composition can exert control on diamond stone size and stone count, and hence diamond grade, as well as geotechnical and processing characteristics.

Diamonds are present in all of the kimberlite units with some variation in grade and stone size distributions. All of the kimberlite material in the A154S, A154N, A418 and A21 pipes is included in the resource and reserve estimates. This is because no mining selectivity is anticipated: all of the kimberlite is expected to be mined.

Early exploration of the Diavik property consisted of airborne geophysical surveys and sampling of glacial till for diamond indicator minerals. Targets were followed up by small diameter core drilling. This work led to the discovery of a number of kimberlite pipes which were further tested by small and large diameter core drilling, described in the next section. Four of the kimberlite pipes were found to have economic concentrations of diamonds and were subject to mini-bulk sampling by large diameter core drilling, underground bulk sampling of two pipes, and feasibility studies leading to the construction of the Diavik Diamond Mine.

The four Diavik kimberlite pipes for which mineral resources and mineral reserves are reported have been delineated and sampled by a combination of small diameter core drill holes, large diameter core drill holes, large diameter reverse circulation drill holes, and Sonic drill holes. Table 3 lists the number of drill holes for each pipe in the period 1994-98 and in 2003-04. The small diameter core is HQ (63.5 mm) and NQ (47.6 mm), fairly standard sizes. In the 1996-98 drilling, the large diameter core was 152 mm diameter (6 in.) with a reduction if required deeper in the hole to PQ size (88 mm). The 2003-04 delineation drilling was HQ size for the deeper holes and smaller BQ core size for shallow holes. The 2004 LDRC holes on the A154N pipe were 24 in. (61 cm) in diameter and on the A418 pipe were 13.75 in (35 cm) LDRC holes and 6 in. (152 mm) Sonic holes.

Figures 7 and 8 show the large diameter core holes and large diameter reverse circulation holes that were drilled on the A154N pipe in plan and section, respectively.

From 1994 to 1998, contacts and geometry of the four kimberlite pipe were delineated by means of inclined holes drilled from surface using HQ and NQ sized core. A number of holes were also drilled from the underground workings. The surface holes were usually drilled radially outwards into the walls of the pipes to define and confirm their circular to elliptical outlines. A total of 19,494 m of drilling was completed in 71 separate holes. This work delineated 104 individual contact points of kimberlite with the host country rocks. The small diameter core hole pierce points were used to construct 3D models of the kimberlite pipes and of the individual kimberlite units comprising the pipes. Small diameter core results were not used for diamond grade estimation.

In 2003 and 2004, additional small diameter core holes were drilled to refine the geometry of the A154N and A154S kimberlite pipes, totalling 4,951 m in 27 holes. They produced about 6 new pierce points for the A154S pipe and about 10 new pierce points for the A154N pipe. The new information resulted in modifications to the geometry of the deeper parts of the kimberlite pipes and a higher confidence level in the kimberlite volumes at depth.

Large diameter core (LDC) holes were drilled in 1996-98 on the four kimberlite pipes for the purpose of obtaining mini-bulk samples to process for diamond content. A total of 11,746 m were drilled in 38 holes. This total includes the PQ core size extensions to the 152 mm (6 in.) core diameter holes listed in Table 3. A total of 266.2 t of kimberlite was obtained in 430 LDC samples which were processed for diamond recovery in a pilot plant. The LDC diamond recovery results were used for grade estimation within the kimberlite pipes.

LDC hole collars were positioned on the ice surface of Lac de Gras at a nominal grid spacing of 15 m to 30 m within the surface projection of the pipe contacts. Holes were spotted closer to the pipe centres than to walls to ensure that they stayed in kimberlite at depth, as is conventional practice, and to avoid intersecting the area of underground workings. All of the LDC holes were drilled vertically to maximize sampling within kimberlite. Since the kimberlite subunits were interpreted to be relatively flat lying, the LDC drilling was considered to obtain representative samples, except possibly near the pipe margins.

The LDC holes reached depths of about 250 m from surface, the nominal depth capacity of the drilling equipment for large core. Deeper downhole core sizes were then reduced to PQ or smaller (4 in. to 3 in. core) to enable sampling to greater depths.

Ten large diameter reverse circulation holes (24 in.) were drilled in the A154N pipe from the 370 m RL bench to provide additional information about diamond grades and values. Four of the holes were drilled deeper in the pipe to a maximum of 381 m and six tested only the top of the pipe.

Five large diameter reverse circulation holes (13.75 in.) tested the upper part of the A418 pipe and four Sonic holes (6 in.) were drilled to delineate the till-kimberlite interface.

For the LDC drill holes, sample intervals were selected to obtain sufficient kimberlite to recover a minimum targeted number of stones or carats for confidence in grade estimation. The sample size thus depends on the grade expected in a particular kimberlite pipe. For the relatively high grade Diavik pipes, DDMI chose nominal sampling intervals of 15 m for 152 mm (6 in.) diameter core and 25 m for PQ and smaller core (3 in. to 4 in. diameter).

Actual sample lengths and weights are somewhat variable within the individual pipes and from pipe to pipe. For the Diavik LDC holes, sample weights varied from 467 kg to 800 kg. Only 3% of the A154S LDC samples had stone densities less than 20 stn/t. Only 16% had lower stone density for the four pipes overall with low values

arising from mud inclusion-rich material, low grade units, or anomalously hard kimberlite with lower process recovery. In RPA's opinion, the LDC drilling provided representative samples.

Overall for the four Diavik pipes, a total of 266 t of LDC material were processed at the DDMI pilot plant in Yellowknife from the four pipes and returned 1,028 ct of diamonds.

The purpose of the Diavik bulk sampling program was to provide a representative sample of diamonds for meaningful valuation and assessment of the average diamond price. A minimum of 5,000 carats was considered to be a representative sample for valuation purposes.

Bulk sampling of the A154S and A418 pipes was carried out by means of a decline and 2 m by 2 m drifts. Pilot holes were drilled in advance of the drifting and allowed grouting and mapping of kimberlite contacts but were not used for sampling. The A154N and A21 pipes were not bulk-sampled underground.

In the A154S pipe, samples were collected from a total of 153 m of drifts excavated at the 260 mRL (150 m depth). One heading trended west-northwest and two sub headings were cut in a northerly direction sub parallel to, and within 20 m of, the eastern pipe wall. Five kimberlite units were sampled in the east half of the pipe.

Three headings, totalling 195 m, were advanced to the northeast and east in the west half of the A418 pipe at 260 mRL. Bulk samples were taken from four separate kimberlite units which were mapped underground.

The kimberlite bulk samples from the A154S and A418 pipes were shipped to and processed in 10 tonne/hour pilot plants, either DDMI's plant at the Con Mine in Yellowknife (DDMI) or at the Ekati Mine Koala site. The bottom screen size was 0.85 mm. The 2,587 t sample from A154S yielded 12,688 ct and the 3,350 t sample from A418 yielded 8,325 ct.

Not all kimberlite units identified in the LDC drilling were bulk sampled in the A154S and A418 pipes, since drifts were restricted to one level and did not extend across the entire pipe. Stone density versus stone size distributions and size probability distributions for LDC and bulk samples, however, compare reasonably well for A154S indicating that the A154S bulk sampling was representative of the diamonds in this pipe.

For the A418 pipe, size distributions indicate that diamonds in the⁺9 sieve size to 3 carats for the bulk sample were proportionately less than for the LDC samples, and that more large diamonds (⁺3 carats) were bulk sampled. This is attributed to under-representation in the bulk sampling of the MFKB kimberlite unit which has larger stones and higher grade than the other units in A418, and to the larger size of the bulk samples which statistically tend to obtain more large stones.

For the 2000 Feasibility Study of the Diavik Diamond Project, diamond valuations were based on stones recovered from the underground bulk samples of the A154S and A418 pipes and on stones recovered from large diameter core samples of the A154N and A21 pipes.

For the A154S and A418 pipes, representative valuation parcels were selected from the diamonds recovered from each bulk sample. The valuation parcels contained all of the stones larger than 3 grainers (0.75 carats) and representative fractions of the smaller size classes. The raw diamond values were adjusted to compensate for under-representation or over-representation of stones in the larger size classes which were modeled to fit a lognormal distribution.

For the A154S pipe, a total of 12,688 carats was recovered from pilot plant processing of 2,587 t of kimberlite at a bottom screen size of 0.85 mm. A representative sample of 5,228 carats was selected for valuation. In the 2000 Feasibility Study, a modeled value of US$79 per carat was used for A154S. In March 2003, Aber sold its first parcel of Diavik diamonds from the A154S pipe for US$96 per carat. Aber estimated that the increased diamond value resulted from improved diamond quality and under recovery of small, low value diamonds.

For the A418 pipe, a total of 8,325 carats was recovered from pilot plant processing of 3,350 t of kimberlite at a bottom screen size of 0.85 mm. A representative sample of 4,717 carats was selected for valuation. The A418 kimberlite was divided into two types and the bulk samples of each were treated separately. The diamond size distributions are different and the two types were valued in separate parcels. In the 2000 Feasibility Study, a modeled value of US$69 per carat was used for the Type A kimberlite and a modeled value of US$49 per carat was used for Type B kimberlite.

For the A154N pipe, since no bulk sample was taken, the 2000 Feasibility Study value of US$33 per carat was based valuation of a relatively small 157 carat parcel of diamonds recovered from mini-bulk sampling by LDC drill holes. A bulk sample was mined from the upper part of the A154N pipe in 2003. A total of 11,771 carats were recovered at a bottom screen size of 1 mm from 19,342 t. These were valued at US$82 per carat.

For the A21 pipe, since no bulk sample was taken, the 2000 Feasibility Study value of US$28 per carat was based valuation of a relatively small 90 carat parcel of diamonds recovered from a 30.5 t mini-bulk sample from LDC drill holes. A larger parcel is needed to confirm the valuation of the A21 diamonds.

Material from individual drift rounds in the A154S pipe was put through the DDMI pilot plant in Yellowknife as two to four-round batches run in two campaigns. Separate batches were run for different kimberlite units. The plant was thoroughly purged between samples to avoid contamination. Some A154S material was also taken to the Ekati Mine and processed in the Koala test plant when the winter road to Yellowknife was closed. The A418 material was stored as piles of aggregated rounds and batched pile by pile at the DDMI Yellowknife plant. The 1996-98 large diameter core mini-bulk samples were also treated at the DDMI Yellowknife pilot plant.

Both pilot plant flow sheets incorporated conventional, proven diamond recovery techniques, and consisted of crushing, scrubbing, dense medium separation, recrushing, x-ray separation, magnetic separation and hand sorting. Bottom screen size was 0.85 mm. Calculated efficiencies for the pilot plants were between 88% and 93% based on recovery and tailings studies, and the use of density tracers in the test work. The pilot plant process flow charts are similar to the Diavik processing plant flow chart.

The initial grouping of recovered diamonds involved combining stones recovered from all of the rounds or piles for each of the three 'geologic units'. These parcels were then sized and split into roughly equal portions by riffling. The 'valuation parcel' included all large goods above 3 grainers (approximately 0.75 carats) in weight and a representative fraction of the small stones. The 'coated' category of goods was recovered on all sieve sizes. Separate parcels were created for coateds larger than the 11DTC sieve size and these went into the representative valuation parcel.

The recrush goods represented stones recovered from the secondary crushing of coarse oversize material reserved from the processing of the various bulk samples. Depending on which test plant was used, either Ekati or the Yellowknife DDMI plant, the recrush was done to either –6 or –8 mm. The recovered goods were combined regardless of the minimum recrush size, and given that only approximately 400 carats of material (mostly lower value small stones) were recovered the overall effect of this material on the average total parcel value is quite small.

The exploration drilling program at Diavik followed conventional kimberlite hosted diamond deposit evaluation practice. Large diameter diamond core drilling (LDC) was employed for mini-bulk sampling of kimberlite. Less expensive smaller diameter coring was used to delineate the kimberlite pipe contacts and provide pipe geometry for resource volume and tonnage determinations. The LDC data were also used for kimberlite geologic modelling, rock density measurements and grade estimation.

At the time of its 2001 due diligence audit for the bank group, RPA reviewed core logging and other data collection procedures and inspected drill logs and drilling records. RPA examined character samples of the kimberlite core, kimberlite hand samples collected from underground and photographs of core. RPA obtained

and examined the resource digital database and conducted checks as part of the audit. The database was audited in 2000 by MRDI and was found to be of very good quality, and essentially free of entry and transcription errors. RPA confirmed these findings. The sample results were digitally integrated into a comprehensive database.

Extensive checking of original logs for the July 2000 update by DDMI resulted in revision of some interval rock formation limits and toe depths for 10 holes. Only one change was significant, the toe of hole 96A154-21 recorded as 250.5 m was revised to 313.4 m. RPA found only one "from-to" entry that was incorrect and this may have been an export/import problem. Several minor data entry errors in a geotechnical file were also found but these have no impact on Mineral Resource estimation.

In RPA's opinion, the drill hole and bulk sampling data are appropriate for estimation of mineral resources and mineral reserves.

BHP-Billiton Diamonds Inc. owns an 80% interest in the Ekati Diamond Mine that is located approximately 30 km northwest of the Diavik Diamond Mine. The Ekati property holdings adjoin the northern boundary of the Diavik mining licences. Ekati began operations in October 1998. Most of the kimberlite pipes mined by Ekati are located some 30 km northwest of the Diavik Diamond Mine, but the Misery kimberlite pipe is being mined a few kilometres north of the Diavik Diamond Mine on the north shore of Lac de Gras.

The Diavik mining leases were originally part of a larger block of claims staked by Aber in late 1991 and early 1992. Under the terms of the Diavik Option Agreement, Kennecott Canada Inc. ("Kennecott") earned a 60% interest in Aber-DDMI block, thereby reducing Aber's interest to 40%. In 1996 and 1997, those mineral claims containing the four kimberlites to be mined along with adjoining claims were legally surveyed and converted to mining leases. In 2001 and 2002, a legal surveying program was completed to convert the remaining mineral claims within the Aber-DDMI block to mining leases. These remaining mineral claims are located to the east and northeast of the Diavik mining leases.

Aber holds interests in several other claim blocks adjoining the Aber-DDMI property. The Aber property acreages and interests are listed in Table 4.

TABLE 4 ABER ADJACENT PROPERTY HOLDINGS
Aber Diamond Corporation — Diavik Mine

The Diavik processing plant has been in operation since November 2002 and is described in a later section.

Tables 5 and 6 summarize the December 31, 2004 mineral reserves and mineral resources, respectively, at the Diavik Diamond Mine, as audited and accepted by RPA. These are as estimated and classified by Diavik, with the exception that RPA has classified the mineral resources of the A21 pipe as indicated mineral resources and inferred mineral resources. The DDMI estimate had all of the A21 mineral resources classified as inferred.

TABLE 5 DIAVIK MINERAL RESERVES DECEMBER 31, 2004
Aber Diamond Corporation — Diavik Mine

TABLE 6 DIAVIK ADDITIONAL MINERAL RESOURCES DECEMBER 31, 2004
Aber Diamond Corporation — Diavik Mine

There are some significant changes from previous reserve and resource statements. The A154N mineral reserve has increased significantly: previous mineral resources below the open pit has been upgraded by more drilling and have been demonstrated to be economic based on a higher diamond value obtained in bulk sampling. The A21 pipe mineral reserves have been declassified to mineral resource because of uncertainty in the economics in the 2004 estimate.

The database for the DDMI estimate of Diavik Diamond Mineral resources is assembled from several sources, as listed below, along with what the type of information is used for.

As noted in previous sections of this report, RPA reviewed some aspects of the Diavik database previously, and we have no reason to believe that there are any material problems with the database. In RPA's opinion, the Diavik database is reasonable and appropriate for mineral resource estimation.

The Diavik mineral resource estimate used simple kriging to interpolate stone density values (stones per tonne stn/t) into a block model using LDC and, for the A154N and A418 pipes, LDRC composited drilling results. The stone density values for each block were multiplied by the average stone weight (ct/stn) for the particular kimberlite unit (domain) to obtain grade values (ct/t). The block volumes were multiplied by the average bulk density for the particular kimberlite domain to obtain block tonnage.

Interpolation of stone density as the principal grade variable is the preferred method where samples are obtained by coring and diamond breakage is minimal. Stone density is less variable than carats per tonne. The latter introduces more nugget effect and less confidence in the interpolation because of the skewed distribution of stone sizes (carats per stone).

Simple kriging weights grade by domain means as well as by the spatial distribution of composites, in contrast to ordinary kriging or other interpolation methods that weight grades only by spatial distribution.

Simple kriging, combined with the relatively few LDC and LDRC samples and the long sample intervals, results in a robust global estimate of resource grade but the grades are smoothed considerably. This means that grades on a local block scale will not predict the in situ grade variability realized during mining. Predicting local grade variability is desirable for base metal and gold deposit mine planning, however, since all of the kimberlite will be mined, mining selectivity is not an issue. In any case, grade variability within the individual modelled units at Diavik is low compared with typical variability in precious and base metal deposits.

After assembly of the mineral resource database, the following procedures were undertaken leading to the block model mineral resource estimate.

The three dimensional solids models of each kimberlite pipe and the kimberlite unit domains within each pipe were modelled from the delineation diamond drill hole pierce points of the kimberlite contacts with wallrocks and other kimberlite units. First, the contacts were interpreted on level plans at regular 30 m intervals, and then the outlines were extruded vertically by computer software. The 3D solids were constructed from contacts on vertical cross sections distributed in a radial pattern according to the delineation drilling.

RPA did some checking of the 3D solids and confirmed the DDMI volumes. In RPA's opinion the Diavik 3D modelling approach is reasonable for resource and reserve estimation.

The LDC samples were composited to regularize the samples for uniform weighting prior to geostatistical interpolation of stones/tonne values. The Diavik samples were composited by equal sample weight rather than the usual practice of equal sample length. The target weight for the composites in each pipe was selected as an average after review of the distribution of sample weights in the pipe. Domain boundaries were ignored and the percentage of each domain was assigned to each composite. This procedure resulted in more composites than original samples, which can lead to bias in interpolation through correlation to a particular geologic unit or grade range. The prior declustering of composites and the kriging process likely addressed any problems in this regard and good agreement of sample statistics with the composites indicates that no bias arose with the compositing.

A considerable body of statistical work has been developed by DDMI on diamond size distributions, stone densities, raw samples, composites, and resource block grades and tonnes both globally and by kimberlite domain. Stone density and size trends with depth in the kimberlites were examined also. Stone density distributions are near normal to slightly skewed.

RPA notes that there are only small differences between the means, the medians, and the coefficients of variation for each kimberlite domain within each pipe and between different pipes for stone densities and size distributions. For the A418 pipe, the bulk sample average grade is 37% lower than the grade for the original LDC samples and composites. This is due to under representation in the bulk samples of the high grade macrocryst-rich kimberlite breccia unit (MFKB) with respect to the LDC samples. Due to the grade contrast of this unit compared to the other A418 kimberlite units, and the difference in diamond values, it was treated as a separate unit for grade estimation and diamond valuation.

Experimental semi-variograms of stone density (stn/t) composites were constructed for each of the four kimberlite pipes to produce isotropic spherical modelled semi-variograms. Variography was carried out on the whole pipe because there were too few composites in individual kimberlite domains to produce meaningful results.

Kriging parameters were derived from the global variography for each kimberlite domain in order to allow for geostatistical interpolation to be constrained to each domain. The variance for each domain was used to define the sill for each domain; in cases where too few samples precluded a meaningful variance, an average

variance value was used. The global nugget to sill ratio was applied to the domain sill to obtain the nugget effect for each domain. The global variogram range for each pipe was applied to each kimberlite domain in the pipe.

From 1994 to 1997, DDMI conducted 3,698 specific gravity (SG) and moisture tests on kimberlite and country rock drill core from both small diameter core delineation holes and large diameter core min-bulk sampling holes. Most of the data are from large diameter core holes. Samples weighing from 100 g to 500 g were collected at 6 m intervals downhole. Testing was done on site and at the Yellowknife bulk sample pilot plant on crushed 1,000 g samples in 1997. Standard water immersion methods were used: the weight-loss-in-water method accounts for 93% of the SG tests with the balance done in 1996 by water displacement.

There were some problems with the 1995 and early 1996 SG determinations and suspect results were discarded or corrected. In 1997 the procedures were modified to splitting the core with one half submitted for moisture testing and the other determined for "wet" SG. The wet SG was corrected for the moisture content of the corresponding split. All SG data were reviewed for reasonableness by DDMI and MRDI as an independent consultant. Samples with suspect SG data were removed from the database and only SG data with identified kimberlite units, as used in the resource solids modelling, were kept in the database. Moisture contents are relatively high in the kimberlite units, and decrease with depth.

RPA reviewed the SG/density data as part of its 2001 audit and noted that the SG testing is industry standard, and procedures are well documented. RPA found that the SG results are reasonable for use in resource and reserve estimation.

Since production commenced in late 2002, much more SG/density data and moisture content data have been collected. In the upper benches mined in the A154N open pit, moisture content has been higher and SG lower than expected. This is because the kimberlite at the top of the pipes, particularly A154S, was not well sampled in the large diameter core drilling. Kimberlite at the top of the pipe contained more water than expected, much of it in the form of ice. Moisture content of kimberlite is in the order of 13% overall average. Also, a mud-rich volcaniclastic unit was encountered that formed a carapace or cap on the A154S pipe. This unit had a lower density than aniticipated. With increasing depth mined, the moisture content is decreasing and the SG is increasing to predicted levels.

For grade interpolation, the block grades were initially interpolated as stone densities (stn/t) and multiplied by the pipe average stone size (ct/stn) to assign a grade in carats per tonne (ct/t) to each block. The interpolation was done by simple kriging, which weights composites within the search ellipse based on the variogram models of grade continuity and for the local mean of the enclosing kimberlite domain. The interpolation for each domain used only composites located within the domain. The search ellipse had a horizontal radius of 100 m and a vertical radius of 50 m.

The mean stone density for the domain, which is first declustered, has a higher influence in determining the block grade where there are few composites. The simple kriging process results is a higher degree of smoothing and places more influence on a robust global average grade for the kimberlite domains, but sacrifices local grade variability somewhat unless composite density is high.

Each block in the model was assigned the volumetric percentages of each kimberlite domain within the block. The average block density for each domain was used to convert the volume of each kimberlite domain into tonnage for each block.

In RPA's opinion the grade interpolation and tonnage estimation procedures are industry standard and are reasonable and appropriate for diamond resource estimation.

Reconciliation studies have been carried out by DDMI to compare tonnes, grade and carats produced from the A154S pipe with reserve estimates in the 2000 Feasibility Study. These reconciliation studies have been complicated by several factors, as noted below, that have contributed to lower than expected diamond grade.

The reconciliation studies accounted for the shortfall in carats produced from the top six benches of the A154S pipe compared to the 2000 Feasibility Study forecast by presence of low grade mud-rich material, volume decrease, lower density and lower plant recovery than predicted. The percentage of carats predicted to carats actually recovered has improved on each bench mined with depth so that it was 95% on the 340 bench.

This reconciliation work led to the use of a 94% metallurgical recovery factor for the processing plant in the production schedule in the 2004 Mine Plan.

At the time of RPA's due diligence audit for the bank group in 2001, RPA reviewed the estimate of Diavik mineral resources and carried out a global check estimate of each pipe using a simple approach based on weighted averages for sample grades from the sample data spreadsheets. DDMI carried out check estimates in 1998 using nearest neighbour, inverse distance squared, unconstrained simple kriging, and constrained and unconstrained ordinary kriging. MRDI audited the resource estimate in 1999 and prepared an independent resource estimate in 2000. Although there were some differences in grade of inferred resources at depth because of fewer samples, the check estimates and audits compared well with the 1998 DDMI mineral resource estimate. RPA's opinion in 2001 was that the DDMI mineral resource estimate was reasonable and acceptable.

Since RPA's 2001 resource audit, the Diavik Diamond Mine has come into operation and has produced significant quantities of diamonds. Production has come primarily from the A154S pipe with some from the A154N pipe.

For the December 31, 2004 resource estimate, DDMI has redone the 2000 block model using the same simple kriging methodology as described above, but with results of more drilling, bulk sampling and operating experience. The lower density at the top of the pipes has been taken into account as has the presence of mud rich volcaniclastic material.

RPA reviewed the 2004 DDMI mineral resource estimate and did an independent block model estimate of the A154N kimberlite pipe, the pipe with the most new data collected in 2003 and 2004. RPA reviewed statistics of various kimberlite units and noted different domains at different elevations within the pipe. RPA used ordinary kriging of ct/t values in LDC drill holes instead of simple kriging of stn/t values used by DDMI. Results of the RPA estimate compared closely with the 2004 DDMI mineral resource estimate for the A154N pipe.

In RPA's view, the DDMI mineral resource estimates for the A154S, A154N, A418 and A21 pipes are reasonable and acceptable.

The key criteria for classification of the Diavik mineral resources are the assessment of geologic continuity, the density of LDC drill hole and sample spacing both vertically and horizontally, and the visual assessment of stone density and stone size profiles from LDC results. The analysis of errors in the estimation of stone density and the results of conditional simulation of the uncertainty in location of pipe wall contacts are also used as a measure of confidence in assigning the classifications for the 2000 estimate.

DDMI classifies its mineral resources and mineral reserves under the Australian JORC code. RPA has reviewed the classification of the Diavik mineral resources with respect to the CIM standards for classification of mineral resources, as required under NI 43-101. RPA agrees with the DDMI mineral resource classifications with the exception of the A21 pipe, as noted below.

The proven mineral reserves in Table 5 are based on measured mineral resources and the probable reserves are based on indicated resources. In RPA's view, there is sufficient geological information are adequate data on and diamond content for classification as measured and indicated resource prior to conversion to mineral reserves.

Mineral reserves are reported for the A154S, A154N and A418 pipes as described in the next section. Only inferred mineral resources below the underground reserves are reported for these three pipes. Additional measured and indicated resources are not reported for these three pipes since most of them have been converted to mineral reserves and the remaining measured and indicated resources are not considered to have sufficient potential for economic extraction to warrant classification as additional mineral resources.

In the 2000 Feasibility Study, the A21 mineral resources were classified as indicated and inferred mineral resources, and the indicated resources in the upper part of the pipe were converted to probable reserves. With the declassification of the A21 reserves to mineral resources, DDMI has reclassified all of the mineral resource as inferred. In RPA's view, it is more appropriate to classify the A21 mineral resources as indicated and inferred under the CIM standards of classification, to reflect level of geological knowledge and confidence in the grade and tonnage estimates.

The following sections describe the basis for the 2004 mineral reserve estimate and the economic justification for designation as mineral reserves. The modifications to the A154S, A154N and A418 open pit designs are relatively minor. The A21 pipe was originally in the mine plan as an open pit and classified as mineral reserves, but is not included in the 2004 mineral reserves because of uncertain economics. The A21 pipe has been declassified to mineral resources, as noted previously. DDMI plans to do more work in future on the A21 pipe to gain more confidence in the mineral resource estimate and diamond value.

Diavik open pit mineral reserves were estimated by a conventional approach. The resource block models for A154S, A154N and A418 were used as the basis for open pit design and optimization using Whittle software. The A154S and A154N kimberlite pipes are both mined in the A154 open pit. The pit designs were finalized with ramp design and adjustment of waste to ore ratios to accommodate the ramp.

Extensive geotechnical work was carried out from 1995 to 1998 to collect data on geothechnical parameters of the kimberlite and country rock. This information was used to establish key operating parameters such as pit slope angles, allowable underground spans and support methods. Inter-ramp angles for pit slopes in the relatively competent granitic pit walls were designed at angles ranging from 49° to 56°.

The A154 open pit design has been reassessed and fine tuned several times based on operating experience since the original design in the 2000 Feasibility Study. The latest revisions were done in May 2004, when the east wall in the A514 pit was moved back to flatten the slopes to the angles recommended in a 2003 Golder study, along with additional smoothing of walls and minor changes in the ramp width.

Modifications were made to the A418 open pit design in 2004. The pit design was adjusted to include a ramp that zigzagged into the pit instead of a spiral ramp that was included in the original design. This change results in significantly less waste being mined in the southeast wall and slightly more waste being mined in the northwest wall. The pit floor was raised from 190 RL to 210 RL.

Bench height is designed at 10 m for all of the open pits. External dilution is estimated at 4%, primarily to allow for inclusion of granitic wallrock in the kimberlite mill feed. Reconciliation studies indicate that 4% is a reasonable estimate. There is no internal dilution since all of the kimberlite will be mined; similarly mining recovery is assumed to be 100%. The concept of a cut-off grade is not applicable since there will be no mining selectivity: all of the kimberlite will be mined to the economic depth of each open pit.

Underground mining is planned to be by underhand cut and fill, cut and fill, and blast hole with fill with access via haulage ramps with sublevel development. Major considerations in mine design are the inherent weakness of the kimberlite and water inflow. In 2005, DDMI is commencing a program of underground exploration and test mining at the A154S, A154N and A418 pipes to test ground and water conditions in order to do more detailed mine planning. In RPA's view, this program is justified and should result in much valuable information.

Mining recovery and dilution estimates are included in the reserve estimate. For the open pits, dilution is estimated to be 4% at zero grade and mining recovery is assumed to be 100%. For the underground mining,

recovery is estimated to be 90% and dilution is estimated to be 11.0% for blasthole stoping, 14.5% for overhand cut and fill, and 16.3% for underhand cut and fill, all at zero grade.

In its determination of mineral reserves, DDMI assess the economics of open pit mining and underground mining. With over two years of operating experience, parameters such as direct and indirect operating costs, ongoing capital requirements, mining productivity, processing plant capacity and recovery, and diamond values. Other parameters, such as underground mining capital and operating costs, dilution and recovery, productivity, ground conditions, water inflow, have been reviewed and reassessed since the time of the 2000 Feasibility Study.

RPA has reviewed the economic studies and parameters used by DDMI to justify the designation of mineral reserves and concurs that they are reasonable and acceptable. The 2004 studies are at the level of at least a prefeasibility study, although they are not titled as such. Later sections describe the mining operations and discuss the long term mine plan. A cash flow model in a later section demonstrates the economics of the mineral reserves.

Table 5 summarizes the Diavik mineral reserves and Table 7 gives a more detailed breakdown of the reserves by underground and open pit. RPA concurs with the mineral reserve estimate and classification into proven and probable categories.

TABLE 7 DIAVIK OPEN PIT AND UNDERGROUND MINERAL RESERVES
DECEMBER 31, 2004
Aber Diamond Corporation — Diavik Mine

Initial mining commenced from the A154S pipe in the A154 pit in 2002. Mine production to the end of 2004 is shown in Table 2. Total tonnes of waste and ore forecast to be mined from 2005 to 2009 are, respectively, 127,970 tonnes and 11,865 tonnes. The average strip ratio is 11.8 to 1 from 2005 to 2009.

Other support equipment includes water trucks, utility vehicles, service truck and sanding truck. All equipment listed is conventional mining equipment and appropriate for the planned application, in RPA's opinion. Four new 91 tonne trucks and one new loader have been ordered for the Diavik Diamond Mine.

Pumping capacity in the A154 open pit was increased when water inflow was higher than anticipated. There is significant water inflow from a northeasterly trending fault that passes through the A154N and A154S pipes.

Ore is currently mined from the A154S and A154N pipes. A separate open pit mine will be developed on the A418 pipe, but first a water retention dike will have to be constructed around the pipe which lies under the water of Lac de Gras. Construction of the A418 dike is scheduled to commence in 2005. In the 2000 Feasibility Study, mining of the A21 pipe was scheduled to be mined later in the project life after construction of another water retention dike. DDMI has determined that the A21 pipe requires further exploration, sampling and analysis to determine if this material can be mined at a profit and hence it has been removed from reserves. More exploration of the A21 pipe is planned.

Underground mining of parts of the A154S, A154N and the A418 pipes below the completed open pits is scheduled to begin in 2008. An underground exploration and test mining program is scheduled to begin in 2005 to further delineate and test underground stoping methods in these pipes. The sequence for mining these pipes, as shown in the 2005 Mine Plan, is shown in Table 9. Mining methods proposed will be tested and modified as more information is gathered from the test mining program. Stoping methods planned for these pipes include underhand cut and fill, cut and fill, and blast hole cut and fill.

The 2005 Mine Plan production schedule is shown in Table 10. Production increases to 2.5 million tonnes of kimberlite per year in 2008, with forecast production of some 10 million carats of rough diamonds. Annual

production of kimberlite decreases to 1.5 million tonnes per year in 2012 as the mine becomes an underground only operation. A total of 26 million tonnes of kimberlite will be mined at an average mine production grade of 3.36 ct/t from 2005 to 2017. A total of 82.2 million carats are forecast to be recovered of which Aber's share is 32.9 million carats.

Year	Tonnes Processed (000s)	Mined Grade (ct/t)	Carats to Process	Plant Recovery	Carats Recovered	Aber's Share of Production (000s of carats)
2005	2,265	4.25	9,626	94%	9,048	3,619
2006	2,300	3.76	8,655	94%	8,135	3,254
2007	2,300	4.75	10,930	94%	10,274	4,110
2008	2,500	3.96	9,889	94%	9,296	3,718
2009	2,500	4.27	10,683	94%	10,042	4,017
2010	2,500	3.32	8,300	94%	7,802	3,121
2011	2,501	2.76	6,900	94%	6,486	2,594
2012	1,525	2.58	3,937	94%	3,700	1,480
2013	1,525	2.53	3,856	94%	3,625	1,450
2014	1,525	2.42	3,697	94%	3,475	1,390
2015	1,525	2.37	3,610	94%	3,394	1,357
2016	1,525	2.41	3,669	94%	3,449	1,380
2017	1,524	2.46	3,747	94%	3,522	1,409
Total	26,015	3.36	87,499	94%	82,249	32,900

Stripping of country rock and glacial till is the major component of the mining at the Diavik Diamond Mine. Production has been scheduled to provide fairly uniform total waste stripping to minimize haulage truck requirements over the life of the mine. The highest waste mining occurs in 2008 when some 28 million tonnes are forecast to be mined. The 2005 Mine Plan includes only Proven and Probable Mineral Reserves.

A processing plant was constructed at the Diavik Diamond Mine site with an annual capacity of 1.5 million tonnes. The plant capacity is being increased to enable processing 2.5 million tonnes of kimberlite per year.

The process used to recover diamonds from the kimberlite consists of primary crushing and scrubbing to remove fine material to tailings, secondary crushing to limit the particle top size to 25 mm, dense medium separation (DMS) to recover heavy minerals including diamonds, re-crushing DMS reject to 6 mm to liberate locked diamonds, wet and dry x-ray sorting plus magnetic separation to recover diamonds from DMS concentrate, single particle x-ray sorting, sizing, packaging and transport to Yellowknife. The plant at the Diavik site is rated at 225 tonnes per hour.

There are two process streams that require disposal. Approximately 40% of the plant feed by weight reports as plus 1 mm minus 6 mm grit. This material is collected in a hopper and trucked to the processed kimberlite containment (PKC) storage area. Approximately 60% of the plant feed is pumped as minus 1 mm slurry to the PKC storage area. The slurry in the plant is thickened to approximately 40% solids and pumped via a three stage pumping circuit. Water from the PKC storage is pumped back to the plant.

Security at a diamond operation is an important consideration. All of the areas where diamonds are produced have very limited access. Card locked doors control access and cameras have been installed in sensitive areas. There are random searches for employees exiting from low-risk areas. High-risk areas such as the recovery plant have 100% search policies. DDMI's security force has a complement of 35 people.

Aber markets its share of production either through Aber International N.V., a 100% owned subsidiary, or through its agreement with Tiffany & Co. Aber reports that demand for rough stones remains robust.

In the 2000 Feasibility Study, as reported in Aber's Annual Information Forms, the diamond prices were US$79 for the A154S pipe, US$33 for the A154N pipe, US$56 for the A418 pipe, and US$28 for the A21 pipe. In March 2003, Aber sold its first parcel of diamonds from the A154S pipe for US$96 per carat. Aber estimated that the increased diamond value resulted from improved diamond quality and under recovery of small, low value diamonds. In July 2003 a bulk sample parcel of 11,771 carats from the top of the A154N pipe was valued at US$82 per carat, which is significantly higher than the earlier valuation of US$33 per carat based on a much smaller sample from LDC drilling.

The following is a list of the major current contracts in place for the Diavik Diamond Mine.

I & D Management Services Ltd. — Mining Manpower Services
Tli Cho Logistics Inc. — Site Services Manpower
Ek' Ati Services Ltd. — Catering and Camp Services
Tli Cho Logistics Inc. — 2004 Winter Road — Fuel Supply
Ek' Ati Services Ltd. — Accommodations Complex Expansion
Denesoline Western Explosives Inc. — Explosives
Imperial Oil — Bulk Diesel Fuel
G&G Expediting Ltd. — Expediting, Off Site Receiving, Freight, Forwarding and Personnel Movement Services
#984228 NWT Ltd. dba SecureCheck — Security Services
Air Tindi Inc. — Air Transport — Passenger and Cargo

DDMI completed a thorough environmental assessment before the site was developed. An environmental staff of nine is responsible for monitoring, directing, reporting and communicating on the environmental issues. DDMI reports that the project is in compliance with all permits, all permits are current and that there are no other environmental liabilities known at this time. The federal and territorial legislation that applies to permits at the Diavik site is listed below. DDMI has implemented the ISO 14001 standard and expected to be certified by December 2004.

Occupational health and safety is managed by a staff of 3, including an on-site contractor that has responsibility for medical treatment and first aid, and for providing occupational health assistance and advice. Total workforce at the Diavik Diamond Mine as of December 2004 was 717 compared to a budget of 684.

For the year 2004, the Lost Time Injury Frequency Rate was 4.0 and the All Incident Frequency Rate was 0.5. The overall health and safety performance at Diavik has been very good compared with industry norms.

Socio-economic activities are managed and reported through DDMI's community and Human Resource departments. Relations with all government and NGO's have been consistently very good.

Aber is subject to taxes in various jurisdictions as a participant in the Diavik joint venture that operates the Diavik Diamond Mine. These include income taxes and mining taxes in the NWT and Canadian federal income taxes. Taxes are included in the cash flow model of the Diavik Diamond Mine, as described in a later section.

The capital cost to build the Diavik Diamond Mine was approximately C$1.3 billion. There are further capital costs forecast for the future, including for construction of the A418 water retention dike and for development of underground mines at the A154S, A154N and A418 pipes. The annual capital cost estimates are included in the 2005 Mine Plan and other documents and are listed on Table 11. The capital costs include an allowance for ongoing reclamation.

The capital costs allows for four additional haul trucks and scheduled replacements for dozers and the plant loader, and light vehicle replacements. Equipment replacements such as pumps, pipes and processing plant items have been allowed for in view of the specific applications and potential abrasive character of the materials. Capital expenses also include activities related to the underground exploration and test mining program starting in 2005, the A418 dike construction, PKC water treatment plant and PKC berm additions. There is no escalation included in these costs.

Operating costs forecast for the Diavik Diamond Mine are shown in the Table 11. These costs are based on a two year production history and estimates of likely future price trends. These are based on the 2005 Mine Plan and comprise direct and indirect operating costs. Direct costs include mining, processing, site services, geology, and administration. Indirect costs include environment, health and safety, human resources, workforce development, community affairs, finance, materials management, and strategic planning. Not included are items such as marketing costs, royalties and Aber's head office expenses. There is no escalation included in these costs.

Year	Capital Costs	Operating Costs
	(C$ millions)	(C$ millions)
2005	180.34	222.39
2006	231.25	221.23
2007	145.27	223.13
2008	95.14	234.33
2009	51.41	291.80
2010	16.21	250.79
2011	31.13	261.86
2012	21.72	237.15
2013	30.63	218.79
2014	16.95	214.78
2015	20.88	211.66
2016	11.71	208.72
2017	16.25	218.67

Total	869	3,015

In order to demonstrate the economic viability of the Diavik mineral reserves, RPA has prepared a cash flow model for the Diavik Diamond Mine (Table 12). The input parameters in the model are as described in the sections above. Production forecast in the model is taken from Table 10; capital and operating costs are taken from Table 11. For diamond values, RPA used US$79 per carat for the A154S pipe production, US$82 for the A154N pipe and US$56 per carat for the A418 pipe. The cash flow model includes only Proven and Probable Mineral Reserves.

RPA has made various assumptions with respect to selling and other costs, which may be different than the assumptions Aber has made in its forward looking disclosure. The cash flow model is presented solely to demonstrate the economic viability of the operation and it is not a forecast of Aber's cash flow from the Diavik Diamond Mine. Two 1% royalties payable to third parties are included in the model. An allowance of 1% of revenue is included in the model to allow for marketing, shipping, handling, sorting and insurance of the final product.

RPA has made certain assumptions in the model, including about taxes. The taxes illustrated in the cash flow model are not intended to reflect the tax position of either Aber or DDMI. For simplicity of presentation, the tax assumptions in the model treat the Diavik Diamond Mine as a standalone taxable entity subject to Canadian and NWT income taxes and mining taxes. This is not actually the case, since each of the Diavik joint venture participants is responsible for its own taxes. Since RPA does not have specific expertise in tax matters, the cash flow model tax treatment should be regarded only as a general approximation. RPA has relied on information available from PricewatehouseCoopers LLP. (PWC) in their summary publications on Canadian Mining Taxation prepared in 2003 and subsequent "Budget Update Summaries" prepared by PWC for 2004 and

2005. RPA cautions that Canadian taxation law is a complex issue and the above cash flow model contains several simplifying assumptions.

RPA has used the cash flow model to demonstrate the sensitivity of the Diavik Diamond Mine to changes in various parameters. The effects on the net present value at a 10% discount rate of varying the grade, the operating cost and the diamond price by 10% are shown in Table 13.

TABLE 12 DIAVIK MINE — CASH FLOW MODEL
Aber Diamond Corporation — Diavik Mine

TABLE 13 SENSITIVITY ANALYSIS (C$ MILLIONS)
Aber Diamond Corporation — Diavik Mine

Parameter	Base Case NPV	+10%	-10%
Grade	1,689.9	1,969.4	1,436.4
Operating Costs	1,689.9	1,589.0	1,781.7
Diamond Prices	1,689.9	1,969.4	1,436.4

Based on RPA's estimates the mine will pay back the capital used to build the mine in 2.5 years from the start of production or mid 2005. The capital cost for construction of the Diavik Diamond Mine was approximately C$1.3 billion and operations are considered to have started in January 2003. The estimate of payback does not include financing charges such as interest.

The 2005 Mine Plan contains detailed plans for the Diavik Diamond Mine from 2005 to 2009. There are other plans on a less formal basis that include mining to 2017, the life chosen for the cash flow model. There are also sufficient reserves available for mining to 2023 but the plans for mining these reserves are not formalized at this point.

RPA has audited the mineral and mineral resource estimates for the Diavik Diamond Mine dated December 31, 2004. In RPA's opinion, the drill hole and sampling database are acceptable for mineral resource estimation. RPA's check estimate of the mineral resources of the A154N pipe compare closely with the DDMI resource estimate. The mining, processing and economic parameters are reasonable and acceptable for conversion of mineral resources to mineral reserves. Reconciliation studies have accounted for the shortfall in diamond carat production compared to that predicted in the 2000 Feasibility Study for the A154S pipe, and the shortfall has been reduced to 5% by the 340 bench. The RPA cash flow model is very robust and clearly shows that the Diavik Diamond Mine is profitable and that the designation as mineral reserves is justified. In RPA's opinion, the Diavik Diamond Mine mineral resources and mineral reserves dated December 31, 2004 are reasonable and acceptable.

Aber Diamond Corporation (2005): — Press release March 9, 2005 — Update Mineral Reserves

Diavik Diamonds Mines Inc. (2004): Joint Venture Management Report, December 2004

Diavik Diamonds Mines Inc. (2004): Evaluation of a North Extension to the A154 Pit Design, July 4, 2004

Diavik Diamonds Mines Inc. (2004): Diavik Joint Venture, 2005 Workplan Budget and 2006 - 2009 Forecast, October 2004.

Diavik Diamonds Mines Inc.(2004) Diavik Strategic Development Project — Slide presentation.

Golder Associates Ltd. (2004): Geotechnical and Hydrogeological Evaluation — Proposed DDMI Exploration Decline, June 24, 2004

Golder Associates Ltd. (2003): Geotechnical and Hydrogeological Investigation, December 30, 2003

MacIntosh Engineering Inc. (2004): Lac De Gras Mine Plan Update August 10, 2004.

MacIntosh Engineering Inc. (2004): Detailed Design Basis Report, August 27, 2004

Nishi-Khon/SNC-Lavalin (2004): Diavik Diamonds Mines Inc. — A154 Dike Extension Study, September 2004

Nishi-Khon/SNC-Lavalin (2004): Diavik Diamonds Mines Inc. — Detailed Design of Dike A418, July, 2004

PricewaterhouseCoopers LLP. (2005) Tax Memo 2005 Federal Budget — February 23, 2005.

PricewaterhouseCoopers LLP (2004): Tax Memo 2004 Federal Budget — March 23, 2004

PricewaterhouseCoopers LLP. (2003): Tax Memo 2003 Federal Budget — February 18, 2003

Stubley Geoscience (1998): Bedrock Geology of the East Island Area, Lac de Gras; prepared for Diavik Diamond Mines Inc., unpublished report, October 1998

This report titled "Diavik Diamond Mine Mineral Reserve and Resource Audit prepared for the Aber Diamond Corporation" dated April 8, 2005 was prepared by and signed by the following authors:

As an author of this report titled "Diavik Diamond Mine Mineral Reserve and Resource Audit prepared for Aber Diamond Corporation" dated April 8, 2005 (The Technical Report), I hereby make the following statements:

	Proven Reserves			Probable Reserves			Proven and Probable
Pipe	M t	Ct/t	M ct	M t	Ct/t	M ct	M t	Ct/t	M ct
A154S	6.4	4.6	29.4	4.1	4.2	17.3	10.5	4.5	46.7
A154N	4.8	2.2	10.7	5.8	1.8	10.3	10.6	2.0	21.0
A418	5.0	3.2	15.9	3.7	3.2	12.1	8.7	3.2	27.9

Total	16.2	3.5	56.0	13.6	2.9	39.7	29.8	3.2	95.6

	Proven Reserves			Probable Reserves			Proven and Probable
Pipe	M t	Ct/t	M ct	M t	Ct/t	M ct	M t	Ct/t	M ct
A154S
Open Pit	6.4	4.6	29.4	1.3	5.0	6.4	7.7	4.7	35.8
Underground	0	0	0	2.8	3.9	10.9	2.8	3.9	10.9
A154N
Open Pit	2.2	2.7	5.8	0	0	0	2.2	2.7	5.8
Underground	2.7	1.8	4.9	5.8	1.8	10.3	8.5	1.8	15.2
A418
Open Pit	4.4	3.2	14.2	0	0	0	4.4	3.2	14.2
Underground	0.5	3.0	1.6	3.7	3.2	12.1	4.3	3.2	13.7

Total	16.2	3.5	56.0	13.6	2.9	39.7	29.8	3.2	95.6

	1994-98 Delineation Drilling		1996-98 Mini-Bulk LDC Drilling		2003-04 Delineation Drilling		2004 Mini-Bulk LDRC and Sonic Drilling
Pipe	1994-98 Delineation Drilling		1996-98 Mini-Bulk LDC Drilling		2003-04 Delineation Drilling		2004 Mini-Bulk LDRC and Sonic Drilling
Pipe	Holes	Metres	Holes	Metres	Holes	Metres	Holes	Metres
A154S	26	7,231	13	4,029	16	2,064
A154N	11	3,407	10	2,837	11	2,887	10	1,479
A418	19	5,210	9	3,319			9	505
A21	15	3,646	6	1,561

Total	71	19,494	38	11,746	27	4,951	19	1,984

Block Name	Block Size	Aber's Interest
Aber-DDMI	393,659.78 acres	40%
Commonwealth	114,692.4 acres	44.4%
Tenby	100,960 acres	35%
Pure Gold North	13,032 acres	20.4%
Pure Gold South	5,059 acres	22%
Westfort	12,819 acres	20.4%
KRL	7,747.5 acres	20.4%

Total	647,969.68 acres

	Proven Reserves			Probable Reserves			Proven and Probable
Pipe	M t	Ct/t	M ct	M t	Ct/t	M ct	M t	Ct/t	M ct
A154S	6.4	4.6	29.4	4.1	4.2	17.3	10.5	4.5	46.7
A154N	4.8	2.2	10.7	5.8	1.8	10.3	10.6	2.0	21.0
A418	5.0	3.2	15.9	3.7	3.2	12.1	8.7	3.2	27.9

Total	16.2	3.5	56.0	13.6	2.9	39.7	29.8	3.2	95.6

		Density and Moisture Content	34
		Block Model Interpolation	34
		Reconciliation of Reserves with Production	35
		Resource Estimate Validation	35
		Classification of Mineral Resources	36
	Mineral Reserve Estimate		36
		General	36
		Open Pit Design and Parameters	37
		Underground Mine Design and Parameters	37
		Economic Considerations	38
	Classification of Mineral Reserves		38
OTHER RELEVANT DATA AND INFORMATION			38
	Mining Operations		38
	Recoverability		40
	Markets — Diamonds		41
	Contracts		42
	Environmental Considerations		42
		General	42
	Health, Safety and Socio-Economic		43
	Taxes		43
	Capital And Operating Cost Estimates		43
		Capital Costs	43
		Operating Costs	44
	Economic Analysis		44
	Payback		45
	Mine Life		46
INTERPRETATION AND CONCLUSIONS			46
RECOMMENDATIONS			46
REFERENCES			46
SIGNATURE PAGE			47
CERTIFICATES OF QUALIFICATIONS			48
		William E. Roscoe	48
		John T. Postle	49

Unit	Size	Number	Comments
Haul Trucks	218 t	9
Excavators	19 m³	2
Front End Loaders	18 m³	1
Tracked Dozer	425 kW	3	leased
Tracked Dozer	302 kW	1	leased
Rubber Tired Dozer	336 kW	1
Grader	200 kW	2
Rotary Drill	250 mm diam	3
Crawler Drill	64 mm diam	1

Mining Area	Activity	Start Year	Notes and comments
A154S	Open Pit	2002	Current mining area
A154N	Open Pit	2003	Current mining area
A418 Dike	Construction	2005	Dike scheduled to be completed by 2008
Underground	Development	2005	UG mining in 2008
A418	Open Pit	2008
A154S	Underground	2008
A154N Upper	Underground	2011
A154N Lower	Underground	2011
A418	Underground	2011
A21 Dike	Exploration

		2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	Total
Production Forecast and Revenue
	Tonnes Processed (000s)	2,265	2,300	2,300	2,500	2,500	2,500	2,501	1,525	1,525	1,525	1,525	1,525	1,524	26,015
	Average Mined Grade (Ct/t)	4.25	3.76	4.75	3.96	4.27	3.32	2.76	2.58	2.53	2.42	2.37	2.41	2.46	3.36
	Plant Recovery (1)	94%	94%	94%	94%	94%	94%	94%	94%	94%	94%	94%	94%	94%
	Carats Recovered (94%)(000s)	9,054	8,145	10,253	9,330	10,057	7,802	6,486	3,700	3,625	3,475	3,394	3,449	3,522	82,292
	Average Diamond Price ($US) (2)	80	80	79	77	71	59	66	67	65	65	66	68	71
	Revenue (US $ Millions)	722	652	812	720	718	458	425	247	235	225	222	234	250	5,919
	Revenue (C$ Millions) at C$1.25 = US$	902	814	1,015	899	898	572	531	309	294	281	278	292	312	7,399
Expenses and Deductions (C$ Millions)
	Total Operating Costs	222	221	223	234	292	251	262	237	219	215	212	209	219	3,015
	Less: Selling expenses (1% of revenue)	9	8	10	9	9	6	5	3	3	3	3	3	3	74
	Less: Royalties (2% of revenue)	18	16	20	18	18	11	11	6	6	6	6	6	6	148
	NWT Income and Mining Tax	2	56	161	139	134	69	54	6	6	6	6	10	11	661
	Federal Income Tax	3	6	56	40	36	16	16	4	5	4	4	7	9	206
	Capital Expenditures	180	231	145	95	51	16	31	22	31	17	21	12	16	869
Cash Flow
	Net cash flow	467	275	399	364	358	203	153	30	24	31	27	46	48	2,425
	Cumulative cash flow	467	742	1,141	1,506	1,863	2,066	2,219	2,249	2,273	2,304	2,331	2,377	2,425
	Net present value at a discount rate of 10%	1,690

Dated at Toronto, Ontario April 8, 2005	WILLAM E. ROSCOE, PH.D., P.ENG. Principal and Consulting Geologist

Dated at Toronto, Ontario April 8, 2005	JOHN T. POSTLE, P.ENG. Consulting Mining Engineer

Dated at Toronto, Ontario
April 8, 2005		William E. Roscoe, Ph.D., P.Eng.

Dated at Toronto, Ontario
April 8, 2005		John T. Postle, P.Eng.