The finding of Gounkoto - gold exploration in Mali
Michael Forrest talks to Joel Holliday, Exploration Manager for West Africa at Randgold Resources Ltd, about gold exploration in Mali.
The existence of gold in West Africa, and in particular Mali, has been known for centuries. In 1433, Emperor Kanku Mussa transported eight tonnes of gold on his pilgrimage to Mecca. Randgold Resources Ltd is a relative newcomer and has been exploring in the region since the 1990s, but has an enviable track record of finding new gold mines using modern exploration techniques. The company has been successful in its greenfield exploration programmes, following a policy of building mines from scratch rather than relying on merger and acquisitions to find the next mine.
West Africa is underlain by a number of cratons of Archean (fewer than 2,500 million years old) and Proterozoic (more than 2,500 million years old) that host greenstone terraines. These assemblages of granites, gniesses, volcanics and sediments often occur in linear belts in ancient continents around the world, and they have been responsible for around 40% of global gold production. However, it was not until the 1980s that large-scale gold finds were made in Mali. One area in particular stood out to Randgold – the western district of Loulo, near the border with Senegal. Here were all the classical components of greenstone gold occurrences. The ground was underlain by Birimian-aged (around 2,100 million years old) metasediments and volcanics in a north-north-west trending belt known as the Kedougou-Kenieba inlier, a triangular outcrop some 250km long. Randgold acquired licences in the area from BHP Mali in 1996, which in turn entered a previous arrangement with the French Syndicat d’Or in 1992, that had discovered the Gara deposit in 1981. Swiftly following the acquisition, Randgold found the Yalea (Loulo) deposit and increased the known gold resources to more than three million ounces. In exploration parlance, this confirmed the inlier as so-called elephant country, ‘an area where large-scale finds have been found with the prospect of more,’ explains Joel Holliday, Exploration Manager for West Africa at Randgold.
The whole of the Loulo area is underlain by Birimian-aged sequences, the most prominent being the Kofi formation consisting of greywacke and a variety of argillaceous, calcareous and tourmalinised sandstones. The lithologies present are interpreted to represent a fore-arc environment with a sequence of older deep water argillites and greywackes to the east of the permit area, and limestone and carbonated clastic sediments, representing a shallow shelf setting, towards the west. From an exploration perspective, however, a critical feature is the extensive laterisation resulting from tropical weathering, and the regolith covering. Overall, only 6% of the area is outcrop. However, several regional linear structures cut across the Loulo area and there exists a strong spatial relationship between gold deposits and regional structures that transgress lower order structures combined with rheologically contrasting lithologies and intrusive bodies. These regional structures strike for more than 50km north to northeast across the permit area (red boxes on the map).
Exploration in this geological and surface weathering environment requires the careful integration of a number of data sources. The blanket of regolith and laterite precludes direct outcropping mapping and masks structures and lithological boundaries. One of the most useful techniques for identifying mineralisation in this terrain is soil geochemistry. Samples from a pre-set depth are taken in a grid over the target areas delineated by the extrapolation of linear features in the landscape and limited outcrop. In the southeast part of the Loulo area, a structural trend was identifi ed by a regional airborne electro-magnetic survey showing a north-north-west trend. This is adjacent to the Senegalo-Malian shear zone (SMSZ), where a number of gold deposits have been found. Exploration within the wider region has established a number of indicator parameters for gold mineralisation, including the association of gold with iron-rich formations identified by aero-magnetic surveys, alteration that exhibited tourmaline minerals and identified in boron geochemistry, and geological mapping identifying host lithologies.
At Gounkoto, a greenfield site with no historic exploration, soil sampling over the structural trend revealed a greater than 10km anomalous zone support by lithological and structural data. GIS analysis generated 10 targets within this zone. Subsequent infi ll soil sampling over the Gounkoto targets returned a two-kilometre long, north-north-west trending plus 30ppb gold in soil anomaly. Initial follow-up work consisted of lithosampling, which returned a number of strongly mineralised results (24.6g/t, 83.8g/t, 48.6g/t and 7.3g/t). These locations were subsequently trenched and sampled, the results of which confirmed the prospectivity of the target. Subsequently two diamond core drill holes were placed, with first hole directly over high values identifi ed in trench samples. The result was 46 metres of gold mineralisation at a grade of 13.6g/t beginning 65 metres below the surface. Another seven diamond core holes were drilled and confi rmed Gounkoto as a significant new discovery.
Although Gounkoto was a new discovery, it bears comparison with other deposits in the Loulo area. Alteration at Gounkoto is complex and similarities are seen with the Yalea deposit 20km to the north, with multiple phases of pink albite and ankerite alteration. Similarities also exist with the Gara deposit (25km to the north) where mineralised quartz-tourmaline is also present. Analysis of the mineralisation indicates that initial metamorphism to greenschist facies followed a strong oxidation and hydration event that particularly affected pelitic and fine-grained units and is marked by the introduction of limonite. Subsequent hydrothermal activity was responsible for silica-albite alteration that affected the silica-rich units. The gold and sulphide mineralisation occurred later during tectonic shearing and hydrothermal events.
Holliday explains, ‘The surface geochemistry initially indicated the mineralisation at Gounkoto’. However the extensive laterites and topography mask the gold signature, making it difficult to distinguish genuine targets from those generated by transported anomalies. In the regional exploration around Gounkoto, a weak soil anomaly was identified in 2002 near to the Millenium Highway that traverses the country.
The trench in which a weak gold anomaly was recognised stopped short of this main road, and its values were not followed up. ‘It was not until 2008 when three rotary air blast drill holes drilled over the anomalies gave results of 0.5g/t, indicating the presences of higher grades,’ says Holliday. The trenching was extended and gave high-grade results and the subsequent two diamond core holes confirmed the discovery. ‘Exploration vehicles had been driving over the discovery site for 12 years,’ continues Holliday.
Randgold fast-tracked the development of Gounkoto from discovery at the end of 2008 to production at the end of 2011. Total resources given in the feasibility study are 5.53 million ounces at a grade of 5.02g/t with the remarkably low discovery cost of US$3.50 per resource ounce against an industry average of around US$40 per ounce. The ore from open pit is treated at the Loulo plant, which by the end of 2011 had produced 138,000 ounces at a total cash cost of US$536 per ounce.
Holliday says, ‘There is no single model for exploration, the geology and surface environment create a unique response to mineralisation and the survey methods developed in the Loulo district have discovered million-ounce resources in a difficult terrane’. Randgold is applying this knowledge to re-define old targets in the region.
Joel Holliday: firstname.lastname@example.org