Proceedings of 10th International Kimberlite Conference

Description, classification and interpretation of kimberlites and related rocks, and communication of that information, underpin the development of three-dimensional geological models used in generating reliable diamond resource estimates. A rationalisation of kimberlite terminology and classification is presented in a practical, systematic framework or scheme. The scheme has five stages and is based on progressively increasing levels of interpretation building upon a series of descriptors that are applied independently of, and prior to, genetic classifications. Stage 1 of the scheme is rock description (alteration, structure, texture, components) and involves only limited genetic interpretation. The components are ascribed to three classes: compound clasts (kimberlitic, mantle, crustal), crystals, in particular olivine, and interstitial matrix (groundmass, interclast cement or clastic matrix). Kimberlitic compound clasts include magmaclasts (e.g. solidified melt-bearing pyroclasts), lithic clasts (e.g. autoliths) and accretionary clasts. Where possible, subsequent stages involve classification and higher levels of interpretation, based on increasing degrees of genetic inference. Stage 2 is the petrogenetic classification into parental magma type and mineralogical type. Stage 3a is the broad textural-genetic classification into coherent kimberlite and volcaniclastic kimberlite. In Stage 3b, coherent kimberlite is further subdivided into intrusive kimberlite or extrusive kimberlite, and volcaniclastic kimberlite into pyroclastic kimberlite, resedimented volcaniclastic kimberlite and epiclastic volcanic kimberlite. Pyroclastic kimberlites can be assigned into two main classes: Kimberley type (formerly tuffisitic kimberlite) and Fort à la Corne-type (formerly pyroclastic kimberlite). Stage 4 incorporates an assessment of the spatial relationship to and the morphology of the kimberlite body from which the rocks under investigation derive. Stage 5 involves more detailed genetic interpretation with more specific classification based on the mode of formation.

The Victor Diamond Mine opened in 2008. Different phases of evaluation were used from 1997 to 2003 to systematically build a resource. The Mineral Resource Model suggested an extremely variable and overall relatively low grade for Victor, but that the diamonds would be of superior value. The development of a robust geological model and understanding the emplacement processes were critical to the design and implementation of sampling programs and to the establishment of a reliable resource for mine planning. Victor comprises two kimberlite pipes; the larger pipe, Victor North, includes the highest predicted grades and is the focus of current open pit mining and this paper. The variability in grade reflects the complex geology of Victor North which comprises two cross-cutting volcanic craters. The later crater was infilled by the products of two separate eruptions; a central high grade pyroclastic kimberlite nested within previous unlithified low grade pyroclastic infill. Extensive mixing produced a wide, gradational, inhomogeneous internal contact zone with variable, but overall intermediate grades. Apart from macrodiamond bulk sampling, the contrasting grade zones can only be differentiated using micropetrography and groundmass mineral compositions which creates a practical mining issue in separating ore from kimberlite waste. Ore versus waste is identified during mining by routine bulk sampling of newly exposed kimberlite which is treated in a separate processing plant. The mining bulk sample data and mined mineral resource performance data show that the Victor North Mineral Resource is accurate, significantly contributing to the success and reliability of commissioning and operating the mine.

The Murowa kimberlite field includes three kimberlite pipes and multiple kimberlite dykes that have been emplaced into the Archaen Chibi granite batholith north of the Limpopo Belt in south-central Zimbabwe. Here we describe the geology of the two largest kimberlite pipes: K1 and K2. Observations of drill core, thin section petrography, and mapping of exposed rocks at the Murowa Diamond Mine are used to describe the internal geology and these data form the basis for three-dimensional geological models of each body. The modeled pipe geometries are combined with observed cross-cutting relationships and textural variations among deposits to interpret the nature of emplacement. K1 is an irregular-shaped, multi-lobed kimberlite pipe occupied by volcaniclastic and coherent kimberlite, and enveloped by kimberlite-poor country-rock granite breccias and volcaniclastic rocks. K1 is interpreted to result from multiple emplacement events, involving kimberlite magmas with varying proportions of gas, liquid and solid phases encountering country rock with different amounts of previous fracturing and brecciation, and intruding/erupting over contrasting timescales. K2 is a steep-sided, sub-circular pipe which flares slightly with depth and is dominantly infilled by massive coherent kimberlite. The textural characteristics of the infill preserved in K2 indicate a less explosive emplacement style relative to K1. Observations of pipe shape, deposits along the pipe margins, and textural modification form the basis for interpreting the relative stage of volcanic development, or ‘maturity’, of K1 compared to some other kimberlites. Designations of pipe maturity acknowledge the uniqueness of kimberlite occurrences, and have implications for interpreting the explosivity and duration of kimberlite eruptions, and for resource modeling of kimberlite pipes.

Combined field observations and numerical modeling provide further context for understanding the development of preferred pathways for kimberlite movement through the lithosphere. Kimberlite occurrences are often observed to be located on brittle structures and clustered along linear trends parallel to known structural trends. Direct structural control by means of magma migration up through the crust along brittle structures is likely in some specific cases, but kimberlite dykes are more often observed intruded into solid rock. We present a case study of the Kimberley district (South Africa) that includes two known kimberlite age populations. By systematically testing different stress orientations in two-dimensional (plan view) numerical models we can determine an equilibrated stress–strain state that was most likely present at the time of emplacement of each kimberlite group. The statistically quantified model results suggest that kimberlites are preferentially located in areas of relatively low mean horizontal stress on a sub-regional scale (approximately 100 km). The models also suggest that structures influence the local stress tensor such that the maximum horizontal compressive stress component may rotate into parallelism with the nearby major structures. The strike of a dyke is preferentially parallel to the maximum horizontal stress, and consequently sub-parallel to the nearby major structures. In this way, the structural architecture of the crust, in combination with regional boundary stress, imparts an indirect control over the intrusion trends. Calibrated structure-stress–strain models can be of practical use in kimberlite exploration, but may also be of use in determining tectonic paleostress conditions.

A field of diamondiferous root to diatreme zone kimberlite intrusions has been discovered on the Rae Isthmus, Nunavut, Canada through a variety of exploration techniques including: glacial sediment sampling, geophysics and prospecting. Regional scale glacial sediment sampling resulted in the discovery of numerous kimberlite indicator mineral-rich samples ultimately leading to ground acquisition, airborne geophysical surveys and kimberlite discovery. A total of eight pipes have been investigated by diamond drilling and an additional eight dykes have been discovered through prospecting. Kimberlite emplacement into the Archean basement rocks of the Rae domain occurred circa 546 Ma. The Qilalugaq kimberlite field occurs within a WNW trending corridor measuring 26 km long and approximately 3 km wide, parallel to significant regional lineaments. Kimberlite intrusions on the eastern edge of the field are defined by a poor diamond tenor and ilmenite-rich dispersions. Westward the till dispersions become more garnet-clinopyroxene-ilmenite dominated with a corresponding increase in the diamond content of the kimberlite sources. Q1-4 is the most significant intrusion and is interpreted as a multi-phase kimberlite pipe having a lobate surface area of approximately 12.5 ha with diamond contents of up to 39 carats per hundred tonnes. The textural varieties of kimberlite present within the Qilalugaq kimberlites include tuffisitic kimberlite breccia (TKB) to hypabyssal kimberlite (HK); country rock dilution is variable and occurs in all units. Kimberlite dykes display variable mineralogy. Kimberlite indicator minerals suggest the mantle sampled by the Qilalugaq field is dominated by a lherzolitic garnet peridotite with subordinate Cr-depleted harzburgite.

The ca. 100 Ma Orion South kimberlite complex consists of volcaniclastic kimberlite deposits which are interstratified with lower Cretaceous continental, marginal marine, and marine sediments. Extensive mini-bulk and bulk sampling with 24-, 36-, and 48-inch diameter drilling and underground mining indicate a ±358 % variation in diamond size and ±127 % variation in diamond grade across the kimberlite complex. Variation is notable both between and within stratigraphically defined kimberlite units. Sorting has played a pivotal role in grain size and abundance variation. These sorting mechanisms resulted in both vertical and lateral grain size and abundance variation of the individual kimberlite constituents. As olivine is the dominant component in the pyroclastic kimberlite and has a similar density to diamond, it is considered to be a good proxy for diamond. This contribution summarizes a semi-quantitative approach of kimberlite constituent data acquisition and analysis that can be utilized as a tool toward assessing diamond potential. Olivine macrocryst size and abundance data indicates a systematic relationship, interpreted as resulting from aerodynamic/hydrodynamic sorting of individual grains during primary volcanic eruption. Sorting is interpreted to have occurred both within the pyroclastic eruptive column and during syn-eruptive transport/deposition. In this contribution we consider the variably pyroclastically sorted deposits of Orion South with respect to the olivine and diamond distributions and the correlation between the two. It can be concluded that pyroclastic sorting processes have played a significant role in the concentration and dilution of olivine macrocrysts and, more importantly, diamonds. This investigation indicates that olivine population characteristics within the Orion South volcaniclastic deposits represent a useful predictor of diamond potential.

The A418 kimberlite pipe, Northwest Territories, Canada, has a typical downward tapering morphology, has been explored to a depth of ~600 m where the pipe has a diameter of ~50 m, and is infilled by volcaniclastic deposits. The pipe-filling volcanic succession has a minimum volume of ~6 × 10⁶ m³ and comprises structurally diverse deposits including finely bedded, variably bedded and massive. The finely bedded volcaniclastic deposits are dominated by surge-like beds containing abundant ash aggregates. Petrographic and geochemical characteristics of the volcaniclastic deposits indicate little contamination by the country rock granite; thus, the pipe excavation phase of the eruption is not recorded by the infill. The lack of preserved within-pipe deposits associated with the excavation of the pipe requires an initial highly energetic explosive eruptive phase that completely cleared the pipe and dispersed material away from the vent. Subsequently, the eruption intensity waned allowing the pipe to progressively infill over time; the character of the pyroclastic deposits requires phreatomagmatic eruptive activity. In the upper reaches of the pipe, a massive poorly sorted deposit cross-cuts and buries the surge deposits with intercalated or gradational contacts between the two facies. The surge beds all dip inwards towards the massive deposit. Repetition of these lithological units at depth indicates a recurring sequence of phreatomagmatic surge deposits cross-cut and overlain by massive pyroclastic or debris-flow deposits, the latter forming during lulls in the phreatomagmatic eruptive activity.

The kimberlite body near Undraldoddi, towards eastern part of Raichur Kimberlite Field of South Indian Diamond Province (SIDP) within the Neoarchaean Eastern Dharwar Craton, has been identified as ‘Tuffisitic kimberlite’. It is primarily composed of macrocrysts/microcrysts of abundant pseudomorphed olivine, minor spinels, rare phlogopite, Cr-diopside and abundant magmaclasts set in a cryptocrystalline chlorite–phlogopite–diopside dominated interclast matrix. The pre-existing crysts and clasts are rimmed by thin mantle of microlites of diopside–phlogopite–titanite which are considered to represent selvages of rapidly cooled and evolved kimberlite-melt. Magmaclasts, with kernels of both olivine (dominant) and xenoliths (mantle and crustal), consist of microcryst/phenocryst of pseudomorphed olivine and rare K-feldspar is set in very fine-grained groundmass of diopside–mica (phlogopite)–spinel–apatite–perovskite comparable to hypabyssal kimberlite. Olivine is completely altered to chlorite-smectite-talc aggregates, phlogopite to chlorite, perovskite to Ca–Ti silicate (titanite) and Al–Mg–Cr spinel to a garnet phase. The development of secondary mineral phases like chlorite–smectite–talc, diopside, titanite and andradite in this kimberlite is the characteristic feature and is probably related to the evolution of ascending kimberlite magma by degassing of CO₂, country rock assimilation at higher level and alterations. Spinels can be classified either as high-chromium spinel (chromite) or as Al–Mg–Cr spinel with a growth rim of Ti-magnetite suggesting kimberlitic trend. Chromite (xenocryst) with Cr₂O₃ ranging from 55 to 59% falls very near to diamond stability field indicating deep source. Some of the xenocrystal phlogopite shows a Group-I kimberlite trend, whereas majority show Group-II kimberlite trend. The whole rock REE patterns of these kimberlites show a Group-I kimberlite trend. The data on whole rock chemistry and individual mineral phase chemistry have been utilised to suggest that a continuum of crystallisation of kimberlite magma was maintained both by gradual increase in silica and H₂O in the residual phase as well as contribution from the country rock. The remarkable similarity recorded between Undraldoddi kimberlite and the diamondiferous tuffisitic kimberlites of Wesselton Mine, South Africa is expected to renew the interest in kimberlite search as well as diamond exploration in this part of the Eastern Dharwar Craton.

A number of late Proterozoic detached basins containing (sub-) greenschist facies sediments overlie the Archaean rocks of the Bhandara Craton. The craton is abutted to the east by the Eastern Ghats Mobile Belt, which is recognised as an accreted terrain with high-grade metamorphic rocks. The study area was covered with stream sediment sampling, airborne hyper-spectral (AHS), airborne magnetic and electromagnetic surveys as part of an integrated exploration programme by De Beers India. Though airborne geophysical data over the Umerkote block produced excellent anomalies with good magnetic dipoles, it was not an effective exploration tool because of the complex geological background, in contrast to the AHS survey. A number of linear positive magnetic features were associated with dolerite dykes in the study area. The multi-disciplinary studies carried out in the area resulted in the discovery of lamproitic bodies. Petrographic and whole rock geochemical analyses were carried out for some of the lamproites which were shown to have magmatic textures and which are extremely altered. The predominance of low chrome (<~62 wt% Cr₂O₃) xenocrystic spinels in the lamproites as well as the stream samples indicate that sampling of spinel-bearing lithologies was outside the diamond stability field. There are a few Harzburgitic garnets reported in the central and southern parts and none from the northern part.

The most striking characteristic of the diamond fields of the Siberian and Slave cratons is the linear distribution of the most significantly diamondiferous kimberlites. The diamondiferous Darnley Bay and Dharma kimberlites and the source area of the Lena West diamonds and kimberlite indicator minerals may lie on a northern extension of the Slave lineament displaced 350 km to the west. Extensions of the Slave lineament to the south may include the Buffalo Hills/Birch Mountain kimberlites of Alberta and the Fort a la Corne kimberlites of Saskatchewan. With its displaced northern and southern extensions the lineament may be 2300 km long. The Siberian diamond fields have a linear distribution over 1000 km with one offset. This lineament has been described as a “Zone of Anomalous Mantle”. A model proposed for the Slave craton in which Paleoproterozoic lithosphere has underplated Mesoarchean lithosphere beneath the Slave lineament and controls the geometry of the lineament is applied to the adjoining northern and southern extensions of the Slave lineament and to the Siberian “Zone of Anomalous Mantle”. It is concluded that there may have been a single “Zone of Anomalous Mantle” formed on the Pacific Ocean side of an Archean craton (Siberia/N. America) in Paleoproterozoic time and the present location of the segments results from major lateral displacements. The Aekit Proterozoic orogenic belt separated, thinned and destroyed the base of the adjoining Siberian and N. American parts of the craton.

Victor Mine is one of 21 known kimberlites within the Attawapiskat kimberlite field, Ontario, Canada. Victor was discovered in 1988 and commercial production began in 2008. In 2008, it was identified that subsequent resource evaluation programs targeting satellite kimberlites within the Attawapiskat kimberlite field required effective integration of multidisciplinary data to identify areas of high prospectivity within the field and identify high-interest pipes early in the exploration pipeline. Systematic relationships were revealed between diamond data and various other data sets (petrography, whole-rock major and trace element compositions, mineral trace element compositions, geophysics, and volcanology) both within the kimberlite field and within individual pipes. The most valuable datasets were identified, gaps in knowledge were determined, and economically relevant projects were formulated. One such project has revealed new emplacement ages for Victor and Uniform kimberlites, which suggest that kimberlite magmas erupted within a relatively narrow time span between ~180 and 170 Ma in the Attawapiskat kimberlite field. The integration of large and complex datasets and the communication between investigators from different fields within the geosciences is highly beneficial to accelerate the discovery-to-production pipeline in the diamond industry.

The Dando-Kwanza concession is situated in the incised highlands of Bié Province in central Angola. An aerial reconnaissance survey conducted in 2005 drew attention to artisanal alluvial mining, in the vicinity of six kimberlite occurrences discovered during colonial times. Exploration commenced in 2007 and the subsequent 3-year work program led to the discovery of 39 new kimberlite occurrences. The concession is underlain by Archean gneisses of the Central Shield Zone of the Angolan Shield, which have been intruded by Paleoproterozoic granitoids at ~1.97 Ga (this study). The basement is overlain by Paleozoic red beds of the West Congolian Inkisi Group which are host to many of the kimberlites. The Lubia cluster kimberlites occur within the northern part of the concession and were emplaced at ~238 Ma, based on zircon U–Pb and mica-in-kelyphite Ar–Ar geochronology. The kimberlites are pipes, some with multiple lobes, with surface areas of up to 16.7 ha. They occur as resedimented volcaniclastic, pyroclastic, and coherent magmatic varieties and have been classified as Group 1 kimberlites. Facies types have been interpreted as crater deposits with narrow hypabyssal feeders and one calcite kimberlite as extrusive lava. Microdiamond sampling of the kimberlites returned low modeled grades and it is postulated that erosion of an enriched and eluviated ferricrete cap on many of the kimberlites may explain the prevalence of alluvial diamonds immediately downstream of the kimberlites. Low diamond grades and the low to moderate interest major element composition of garnets and chromites may be related to perturbation of the SCLM by high-T melt-related processes during Paleoproterozoic times.

The K2 kimberlite is part of the Koidu Mine, located on the Man Craton in Sierra Leone, West Africa. The kimberlite is Jurassic in age, 0.5 ha in size at surface, and was emplaced along a pre-existing dyke system into extensively brecciated and locally altered granite country rock. The pipe is infilled by variably sorted, inhomogeneous volcaniclastic kimberlite (VK) and by less common coherent kimberlite (CK). A resource evaluation programme was conducted from 2008 to 2010 to advance the mine plan. This involved drill core logging, indicator mineral abundance and composition studies, bulk sampling, surface mapping, targeted mining, three-dimensional geology modelling and diamond grade estimation. The conduit-filling VK deposits are interpreted to represent several explosive eruptions. The internally complex and variable distribution of the VK rock types reveal both concentric deposition from the pipe margin towards the centre, and vertical stratification creating sub-horizontal domains. Post-pipe emplacement CK intrusions are common and frequently show mixing between the intruding kimberlite and the poorly to unconsolidated VK infill. Secondary alteration (serpentine, clay) has modified the kimberlite mineralogically and to a lesser degree texturally. The K2 pipe infill is characterised by many features consistent with the historical kimberlite classification ‘tuffisitic kimberlite’ (TK), but is more highly variable than typical large ‘TK’-infilled bodies. Diamond grade and value information obtained from large diameter drilling and targeted mining were integrated with simplified geological models to derive global estimates of diamond grade and value for geologically defined resource domains. These formed the basis of a feasibility study of the Koidu kimberlites and positive results led to expansion of the mining project.

Sixty-two kimberlites discovered on Hall Peninsula, southern Baffin Island, Nunavut between 2008 and September 2011 comprise Chidliak, the most recently discovered kimberlite province in Canada. These discoveries were the result of traditional kimberlite exploration techniques used in glaciated terrains, including kimberlite indicator mineral (KIM) sampling and analysis, ground and airborne magnetic and electromagnetic geophysical surveys, and ground prospecting and drilling. Capacitively coupled resistivity ground surveys and comprehensive KIM classification and interpretation techniques also played a significant role. Both sheet-like and larger pipe-like bodies have been discovered. The sheet-like bodies are mainly steeply dipping hypabyssal kimberlites, which may contain basement xenoliths. The majority of the pipe-like bodies have pyroclastic or apparently coherent kimberlite infill containing sedimentary xenoliths derived from now-eroded Paleozoic strata in addition to basement xenoliths. The presence of these sedimentary xenoliths, along with other textural features, suggests that at least some of the apparently coherent kimberlites are not intrusive but are extrusive, either effusive or clastogenic in origin. Many of the kimberlites manifest as strong remanently magnetized 0.5–2.0 ha bodies that are expressed as magnetic high and low anomalies with associated weak and shallow conductivity responses. Bodies dominated by pyroclastic kimberlite infill can have neutral to weak magnetic signatures, appear to be more conductive, and can be much larger (up to 4–5 ha) than those with dominantly coherent kimberlite infill. Perovskite U–Pb dating of 25 of the kimberlites indicates magmatism spanned a period of approximately 18 million years, from 156 to 138 Ma (Late Jurassic to Early Cretaceous).

The Archean Dharwar craton of India hosts well-known kimberlite fields. The present study area around Narayanpet–Maddur in Andhra Pradesh of South India is a part of the eastern Dharwar craton with ~34 known kimberlite bodies (1400–1080 Ma) that form three clusters. Geophysical studies help to decipher the tectonically complex nature of the area. Regional gravity and magnetic anomaly maps show the subsurface structures in the area depending on the density and susceptibility contrasts respectively. The eastern and western parts of the study area are comprised biotite granite and migmatite gneisses that are separated by a prominent NW–SE trending schist belt that is reflected in both the gravity and magnetic data. Based on the geophysical data, this belt is likely to be the result of basement faulting. Other NW–SE trending deep-seated faults are clearly delineated in both geophysical maps. Along with NW–SE trending major faults, E–W, NE–SW and N–S linear trends indicate structural disturbances, and the intersections of lineaments are likely favorable locations for the emplacement of the kimberlites. Magnetic anomalies are more effective in the identification of faults, fractures and dykes in the granitic terrains. Application of various digital filters to the geophysical data clearly indicates that anomalies correlate with the known surface geology and provide information about the shallow and deeper subsurface features. The analytical signal map demarcates the weathered and unweathered pipes based on the intensity of magnetization.Gravity and magnetic 2D forward modeling of profiles across the Narayanpet, Maddur and Kotakonda clusters reveals intrusive bodies in the host rock and the structural features in the area. Both models correlate with each other in terms of the subsurface geology and the depth of intrusive bodies. Euler deconvolution depth estimation for gravity and magnetic data brought out the location and depth of dykes as well as new potential kimberlite areas to the north of Maddur and in the NE and SW portions of schist belt. These potential areas lie within a conjugate system of fractures and faults, some of which may have been reactivated.

Renard 3 is one of a ~640 Ma. cluster of nine diamondiferous kimberlite pipes in the Otish Mountains region of Québec, Canada. The external and internal geology of Renard 3 were investigated to assess the resource volumes and diamond contents, information that was used to develop an Indicated Mineral Resource of 106 carats per hundred tonnes. The internal geology was determined by differentiating distinct kimberlite units using a combination of megascopic, macroscopic, and microscopic features. The units were assessed and modelled as five distinct, vertically extensive pipe-forming phases of kimberlite, each separated by steep and predominantly sharp internal contacts. Each phase is dominated by one of the following textural varieties of kimberlite: (i) volcaniclastic kimberlite that is classified as tuffisitic kimberlite breccia, (ii) coherent kimberlite that is classified as hypabyssal kimberlite, or (iii) transitional kimberlite that is intermediate between tuffisitic and hypabyssal. The combination of observed kimberlite textures, rock types, associated country rock features, emplacement age and regional geological setting indicates that Renard 3 is an eroded, deep diatreme to root zone kimberlite pipe comparable to others in the Renard cluster, Gahcho Kué, Northwest Territories, Canada and Kimberley, South Africa. The worldwide geological correlations significantly increase the degree of confidence in the Renard 3 geological model and, in turn, the Mineral Resource Estimate, highlighting the importance kimberlite geology has in the economic assessment of a kimberlite.

The North Australian Craton occupies a large part of Australia’s Northern Territory. However, while much of the Territory is prospective for diamonds, its large size has meant that it is under-explored. Historical exploration methods have been reviewed and critically assessed. Combined with data on mantle lithosphere thickness and structure, and the ages of primary diamond deposits compared to regional rocks, results have been used to rank and model regional diamond prospectivity. The ranking scheme has been supplemented with qualitative modifications and it is concluded that the Territory’s western and eastern flanks are the most prospective. The Territory’s three known diamondiferous kimberlite fields were all discovered using indicator minerals as the principal exploration technique. However, due to intense Cenozoic weathering, traditional indicators are unusually small or absent. Surveys using coarse-spaced sampling or that ignored fine sand particles (<0.25 mm) may therefore have returned false negatives. Indicators are often outnumbered in exploration samples by microdiamonds, considered as unreliable as regional prospecting vectors due to wind transport and their ubiquity in the east. Alteration resulting in elevated Ti relative to Mg in ilmenite and Zn overprints on chromites, and relatively high G9-garnet abundances in diamondiferous pipes place many diamond indicator compositions outside preferred discriminatory fields. The effects of intense weathering on traditional indicator mineral compositions are expected to promote the use of more durable indicator minerals in future exploration, and place a heavier reliance on remote sensing. Given the large areas involved, future diamond exploration will most effectively be conducted as parallel programs to established exploration for other commodities.

As a result of multiple issues, alluvial diamond properties are often considered more difficult to evaluate than many other deposit types. This is frequently combined with the perception that only large-scale deposits can be evaluated professionally; that evaluations and, subsequent, valuations of small-medium scale operations are either inherently flawed or cannot be done at all. However, a discussion of resource estimation and valuation methods shows that, despite both real and perceived risks, alluvial diamond projects of all scales can be evaluated to the same international standards in a financially responsible manner. Once the site has been established, bulk-sampling and drilling programmes can be designed to estimate the diamond resources present on the property. It has to be appreciated that, due to the often low, but highly variable, grades and geological inhomogeneity of these deposits, significantly more prospecting may need to be completed to estimate resources when compared with many other commodities, as unsubstantiated assumptions can be misleading. The resource estimation programme is designed to identify the volume of gravel present on the property that may be mineable, the average recoverable grade and the average sales value of the diamonds. This can only be accomplished by the processing of significant volumes of gravel through bulk-sampling. During trial-mining exercises, which are comparable to (pre-) feasibility studies, considerable mining and processing data is obtained—the “modifying factors” required to convert mineral resources to mineral reserves. As with any other commodity, the value of an alluvial diamond deposit changes as market conditions change and according to the amount of information available for the project. Consequently, different valuation approaches will be valid at different stages of the project. Where reserves have been identified, standard Discounted Cash Flow (“DCF”) methods of valuation apply. However, few alluvial diamond deposits get evaluated to this stage and Income-based approaches are not appropriate for properties without mineral resources. By contrast, Cost and Market approaches, which are often used in early stage project valuation, are generally not good indicators of value of alluvial diamond projects either.

De Beers operates two diamond mines in Canada, the Snap Lake Mine, Northwest Territories and the Victor Mine, Ontario. Because mining involves the extraction of finite resources, the sustainability of a mine can be defined by the extent to which it can serve as a catalyst for benefits that extend well beyond the life of mine. The challenge in mining, therefore, is to work with affected communities to utilise the value realised by a mining operation to assist affected communities to realise their long-term goals and aspirations. Both the Snap Lake and Victor mines commenced production in 2008, and each mine affects four separate communities. Many of these communities are remote fly-in villages, with seasonal road access for less than 2 months a year; one village has access only by rail or by air, while the remaining three communities are accessible year round by road. De Beers’ experience is described to date in terms of approach, successes, failures and challenges. De Beers’ has seven impact benefit agreements (IBA) with eight main communities of interest (two communities have a single IBA): these agreements are the foundation of company-community relationships. Important aspects include the interaction of company personnel and community members, frequency and duration of visits to communities and the mine sites, together with availability of adequate resources for both company and communities. Significant progress has been made with employment, training including education, and business opportunity initiatives, as well as cultural enhancement. Furthermore, the focus has been on those skills and opportunities that are not specific and unique to mining, so that there will be a wider range of future opportunities for those who participate. Implementation of agreements is more complex and requires more effort than planned, and even after 5 years of implementation, community capacity issues are constraining. Proponents should keep these facts in mind for future projects.