Geology & Resources

Regional Geology

Geological map of Sierra Leone showing the location of the Koidu Kimberlite Project.
Geological map of Sierra Leone showing the location of the Koidu Kimberlite Project.
Click to expand | Download PDF [2.0MB]
Regional Geological Setting 
of the Koidu Kimberlite Project
Regional Geological Setting of the Koidu Kimberlite Project
Click to expand | Download PDF [2.8MB]
Schematic representation of the shear fractures developed during dyke emplacement, the dyke geometry and the controlling stresses (Barnett, 2009).
Schematic representation of the shear fractures developed during dyke emplacement, the dyke geometry and the controlling stresses
(Barnett, 2009).

Sierra Leone is situated on the Man Craton of the Southern West African Shield. The Archean granitic shield contains elements of early sedimentary and mafic formations and a group of supracrustal greenstone belts with banded ironstone and detrital sediments. The basement granite and rocks of the younger Kambui Group have been deformed and metamorphosed together with the underlying Leonean gneisses and intruded by late and post orogenic granites.

The granitic rocks are cut by several fracture systems which are widely believed to have controlled the emplacement of kimberlite and dolerite dykes. Foliations and faults in the basement granites are almost parallel, trending roughly north-south to northeast-southwest. Second order fractures developed in northeast orientations. Vertical amphibolite dykes in the Koidu area, formed by the metamorphism of dolerite (diabase) dykes, are sub-concordant with the host gneissic basement. The dolerite dykes and sills of Jurassic or Cretaceous age, generally have strikes of between 110° and 130° (Hall, 1968).

Three dominant brittle structural patterns were observed by Williams (1977) in the Archean gneisses, which are N-S, ENE-WSW and NW-SW, and these were confirmed by Barnett (2009) during a detailed structural assessment of the Koidu Kimberlite Project. Barnett (2009) also notes that both Williams (1979) and Venkatakrishnan and Culver (1989) suggest that the N-S and ENE-WSW brittle structures form as conjugate shears in response to the N-E directed maximum compressional stresses caused by Neoproterozoic orogenesis, specifically the Rokelide orogeny in western Sierra Leone.

Venkatakrishnan and Culver (1989) suggest, based on an oblique rifting model, that the dominant reactivated structures during the Mesozoic rifting would be approximately N-S to NNW-SSE trending normal faults reactivated along ENE-WSW trending sinistral strike-slip corridors.

The Koidu and Tongo kimberlite clusters of pipes and dyke zones are part of a Jurassic age province of kimberlites within the Man Craton, which extends from the 154 Ma Droubja kimberlite in southeast Guinea and curves around to the +145 Ma bodies at Koidu, then extends southwards to the + 140 Ma bodies at Tongo. The distribution of the kimberlites is thought to be related to continental scale movements and tectonic stresses established in the lithosphere rather than to local scale structures (Skinner et al., 2004). Skinner et al. (2004) presented evidence in support of hotspot activity being related to kimberlite genesis and, assuming that kimberlites are generated from a fixed position in the asthenosphere, that it is possible to track the direction and rates of movement of the overriding lithosphere from known kimberlite distributions and ages.

Kimberlite Geology

The main pipes, named K1 and K2 are smooth, steep sided bodies that are morphologically similar to those mined in the Kimberley area of South Africa (Class 1). Present surface expressions of the pipes are approximately 0.45 ha for K1 and 0.50 ha for K2. The external morphology and infill present within the pipes is consistent with a lower diatreme setting and significant erosion of the pipes has occurred. The pipes are infilled by multiple phases of kimberlite characterised by contrasting textures due to different emplacement processes (highly explosive vs. intrusive). Texturally, the infill within the bodies is dominated by massive to locally bedded volcaniclastic kimberlite classified as tuffisitic kimberlite (TK). The volcaniclastic rocks typically contain a high proportion (15-90%) of fresh granite xenoliths and variable proportions of olivine, mantle derived mantle-derived xenoliths and xenocrysts (the latter are referred to below as indicator minerals).

Coherent kimberlite (CK) is less common but volumetrically significant within the pipes and occurs as main pipe infill, as well as late stage dykes and rare sills (classified as hypabyssal kimberlite – HK). The different rock types or phases of kimberlite present within the pipes are characterised by different grades.

In addition to the well-formed pipes, there are a number of blows that represent poorly developed, small, volcanically-immature pipes. These bodies are named Blow A, Blow B1, Blow B2 and Blow B3. These bodies are characterised by more complicated external pipe shapes compared to K1 and K2 and are dominantly infilled with CK, texturally transitional kimberlite (characterised by both coherent and volcaniclastic features) and less common, well developed, massive volcaniclastic kimberlite classified as TK.

The four main dyke zones (DZ), termed DZA, DZB, DZC and DZD, were emplaced along a southwest to northeast structural trend and are classified as Group 1, macrocrystic, phlogopite (+/- calcite and monticellite) HK. Structurally the dykes consist of irregular, braided and en-echelon arrays typically made up of multiple segments each ranging in thickness from a few cm to just over 3m. Simple single segment dykes are less common. In addition to variations in external morphology, the dykes display internal variation in the size and proportion of olivine macrocrysts, the type and abundance of mantle derived indicator minerals, mantle xenoliths and diamond grade.

Associated with the main pipes and blows, and less commonly with the dykes, are locally extensive zones of leached granite (SiO2 removed) and marginal (or contact) breccias that typically contain low proportions (<10%) of kimberlite. These zones are interpreted to both predate and postdate the formation of the pipes and have been incorporated into the geological models. Concentric ‘onion-skin’ shells of fractured and altered granite surrounding rounded granite cores are also observed, as well as joints filled with pulverised, angular shards of country rock around many of the dyke exposures and pipe walls. Although the marginal breccias, leached granite and onion skin granite zones in most cases are not of economic interest, they represent geological weak zones that will impact mining activities.

K1 Kimberlite Pipe

K1 is a steep-sided but slightly irregularly shaped pipe emplaced along Dyke Zone A and Dyke Zone B, and closely associated with Blow B1 and Blow A. It has an estimated surface expression of 0.45 hectares and has been modelled from the present surface (current pit bottom at ~270 m amsl) to -170 m amsl. K1 is currently open at depth.

The pipe is occupied by three main kimberlite types including a volumetrically significant coherent kimberlite infill (KIMB1) and two varieties of tuffisitic kimberlite, KIMB2 and KIMB3, the latter being volumetrically relatively minor. KIMB1 and KIMB2 have each been subdivided into two subtypes based on xenolith content (KIMB1A, KIMB1B, KIMB2 and KIMB2B). Several late stage hypabyssal kimberlite dykes have intruded K1, primarily within KIMB2 and KIMB3.

The geology model generated for K1 comprises 5 internal kimberlite domains (one for each main kimberlite infill type) and a number of external pipe solids encompassing zones of marginal breccia and leached granitoid as intersected by drilling.

Resource estimates for K1 are based on the geological and bulk density data from 31 core drill holes and production records from recent mining history. Production records used for resource estimation represent ~ 295,000 tonnes of kimberlite mined between 2005 and 2007, and processed by the Koidu diamond recovery plant to yield ~ 196,000 carats of diamonds.

Based on these data a total indicated resource of 1.38 million tonnes has been declared, encompassing two separate resource domains (one comprising KIMB1 the other KIMB2 and KIMB3) that together make up the entire pipe down to a level of 150 m amsl. The estimated diamond grade for this resource is 0.67 ct/t. In addition to the indicated resource an inferred resource of 3.56 million tonnes is declared for the portion of K1 below 150 m amsl, extending to the base of the current pipe model at -170 m amsl. The estimated grade and average diamond value for this material are the same as for the indicated resource.

K2 Kimberlite Pipe

K2 is a steep-sided, slightly southward plunging pipe emplaced along Dyke Zone A. It has an estimated present surface expression of 0.50 hectares and has been modelled from the present surface (current pit bottom at ~315 m amsl) to -257 m amsl.

The K2 pipe is internally complex. It is dominated by a single main pipe infill (KIMB1), a highly variable tuffisitic kimberlite that has been further subdivided into seven sub-units characterised by varying olivine grain sizes and proportions of xenolithic country rock and/or juvenile material. Numerous additional, volumetrically minor volcaniclastic pipe infills are observed, including two types of tuffisitic kimberlite and a transitional kimberlite. The volcaniclastic infills have been extensively intruded by late stage hypabyssal kimberlite material, including multiple dykes and two volumetrically significant intrusions.

The geological model constructed for K2 comprises eight internal kimberlite domains and two external domains (encompassing marginal breccia and leached granite). The pipe can be broadly subdivided into a central core dominated by different varieties of KIMB1 and a marginal zone comprising a mixture of KIMB4 (modelled separately) and a mixture of fine grained, highly diluted TK (KIMB1Bm) and common intrusive HK (modelled together as KIMB1-MZ).

A significant amount of evaluation work has been undertaken on the Koidu K2 pipe over an approximately two year period. Factoring in historical drilling work, a total of 35 core drill holes, eleven 17 inch LDD holes and 13 large surface bulk (BSU) samples were undertaken. In addition to the geological data provided from these sources, a total of 81 well distributed bulk samples with a total volume of 485 m3 (~ 1,200 tonnes) were obtained from the LDD drillholes for grade estimation purposes, and in excess of 21,944 m3 (~54,000 tonnes) of kimberlite was excavated by surface sampling to provide a large parcel of diamonds for estimation of average diamond value. Bulk samples were all processed by the Koidu diamond recovery plant using a bottom cut-off of 1.2 mm. The total amount of diamonds recovered during this evaluation exercise includes 459 carats from LDD sampling and 21,588 carats from the BSU samples.

Based on this evaluation work, resource estimates have been compiled for the K2 pipe. A total indicated resource of 2.79 million tonnes has been estimated, encompassing 5 separate resource domains within the volumetrically dominant central portion of the pipe down to a level of -30 m amsl. The estimated diamond grade of this resource is 0.34 ct/t. In addition to the indicated resource, an inferred resource (covering marginal pipe material to a depth of -30 m amsl) of 1.62 million tonnes, with an average diamond grade of 0.32 ct/t is estimated. An additional approximately 2.71 million tonnes of kimberlite has been estimated in the deeper portions of K2 (from -30 m amsl to -257 m amsl) but is too poorly constrained to be included in the present resource estimate.

Kimberlite Dyke Zones

DZA and DZB are complex dyke zones comprising multiple sheets (on average ~3 to 6 sheets per drill intersection) whereas most intersections of DZC and DZD only comprise a single dyke sheet. For all dykes segments three-dimensional surfaces were modelled (through the centre point of dyke intersections) as a basis for resource estimation work. Additional modelled solids were generated for each dyke zone to encompass  the main dyke intersection and the entire dyke zone, respectively. The confidence level of dyke zone models is highly variable and in places low due to a paucity of drill intersections and a lack of surface mapping data.

Petrographic investigation indicates that there is relatively insignificant variability at the thin section scale within each dyke zone and among dyke zones (although mineralogical variation does exist between certain dyke zones). On the basis of this investigation, each dyke zone is interpreted to comprise a single geology domain which shares common features. However, that macroscopic textures (i.e. as observed in hand samples, field outcrops, pit exposures) suggest there may be significant internal variability with respect to the olivine content and internal dilution. Such variability may have an impact on the distribution of grade within the dyke zones.

Where possible, the volume, tonnage, grade and average diamond value of each dyke zone segment has been estimated based on a combination of the results of geological modelling  and data from mining and bulk sampling of DZB and DZA, respectively. The data are considered appropriate for estimation of an inferred resource for the dyke segments DZA-1, DZA-2, DZB-E and DZB-W of 4.27 million tonnes at an average grade of 0.54 ct/t. Due to an absence of data constraining the grade of DZC, no resource is defined for this segment, however tonnage estimates are provided for this body, which is considered to represent geological potential. Due to a lack of data, no resource or volume estimates are provided for DZD at present.

Kimberlite Blows

The four kimberlite blows at Koidu are all steep-sided, volumetrically relatively minor, pipe-shaped bodies. With the exception of Blow B2, multiple rock types are evident in all of the blows. Between 2 and 4 kimberlite infill types are defined for each blow, including both volcaniclastic and coherent kimberlite types. However, due to the small size of these bodies, the fact that in most case a single infill type is volumetrically dominant, the limited number of drillholes on most of the blows and the preliminary nature of the geology review, no attempt was made to model separate internal kimberlite units and each blows is modelled as a single domain.

Resource estimation work was undertaken on Koidu Blows A, B1 and B3 based on data derived from a combination of drillhole intersections and surface excavation, and processing of significant volumes of kimberlite. No estimates have been generated for Blow B2 which is very small and, at the current level of exposure can reasonably be considered as a minor enlargement on Dyke Zone B.

Geology and bulk density data used for volume and tonnage estimates for Blows A, B1 and B3 are derived from a total of 36 core drillholes. Grade data for these bodies derived from mining and bulk sampling that yielded a total of ~114,000 tonnes of kimberlite processed by the Koidu diamond recovery plant to produce a total of ~ 72,000 carats. Based on these data, an indicated resource of 0.042 million tonnes containing 0.032 million carats of diamonds has been estimated for Blow B3 and an inferred resource of 0.071 million tonnes containing 0.038 million carats of diamonds has been estimated for Blow A. The grade of Blow B1 is not sufficiently well constrained to permit estimation of a resource for this body which is estimated to contain 0.517 million tonnes of kimberlite classified as geological potential.