Sediment-hosted gold is hard to explore for because the gold does not follow the hydrothermal rules most of us learn first. There is no obvious alteration halo or vein array, and the deposit often has no surface expression. The Witwatersrand Basin in South Africa is the best-documented example. Estimates vary with the source and the method. One puts it at 30–40% of all the gold ever mined, over 1.6 billion ounces, while another gives about 22% of the gold accounted for above the surface. By the end of this article you should be able to read a sedimentary basin as a gold-distribution machine. That means identifying the likely source, the transport path, the trap and the later overprint, and turning those into a ranked drilling sequence.

The underlying principle: gold behaves like a heavy grain

Start with plain physics. Gold is about 19 times denser than water, and far denser than quartz. In a moving river, light grains travel on while gold drops out wherever the current loses energy or finds a trap. Traps include the base of a gravel bar, a scoured bedrock surface and the lee of a pebble. Repeat this over millions of years and you get a placer, a sedimentary concentration of heavy minerals.

The Witwatersrand adds three technical layers to this.

• A basin that kept receiving sediment. The Central Rand Group was deposited in a retroarc foreland basin that regressed as it filled, between roughly 2.90 and 2.78 Ga. A foreland basin is a depression formed by the load of an adjacent mountain belt. Important gold deposits formed on braided fluvial plains that sometimes ended at a shoreline.
• A source upstream. The most likely source of the detrital gold is older lode gold deposits in the greenstone belts of the Kaapvaal Craton.
• Pauses in deposition. Tectonic uplift triggered most conglomerate reefs, while long stable periods let low-energy processes and unconformities winnow heavy minerals. An unconformity is a surface representing a gap in the rock record.

The accepted interpretation is the modified placer model. It integrates placer and hydrothermal ideas. Gold arrived as detritus, then was redistributed on a small scale after burial. Brittle deformation linked to the Vredefort impact at 2.02 Ga is one proposed trigger for late remobilisation, and one microbeam study found remobilisation over distances of less than centimetres. The debate is not closed, since a hydrothermal school argues the gold was introduced after the host rocks formed. For targeting, the practical point is the same either way: sedimentary architecture controls where the gold is.

A workflow for sediment-hosted targeting

1. Define the basin and its margin. Use geological maps, SRTM elevation data (about 30 m resolution) and regional gravity and magnetic grids to find the basin edge and any buried structure. The aim is to locate the basin-margin where coarse sediment entered.
2. Identify candidate source terranes. Look for gold-bearing greenstone or other older mineralised rocks upstream. Check detrital zircon ages if published, because they link sediment to source.
3. Reconstruct palaeo-flow. Collect or digitise cross-bedding directions, pebble-size trends and channel orientations. Coarse gravel tends to sit near entry points. Heavy minerals such as gold and uranium are concentrated in coarse fan-delta gravels, although the favoured position within a fan differs between reefs.
4. Map unconformities and reworked horizons. Use borehole logs and outcrop to find the surfaces where winnowing occurred. Gold in the Ventersdorp Contact Reef mainly occurs where it unconformably overlies reefs of the Central Rand Group. Reworked older placers can feed younger ones, so check what lies directly beneath an unconformity.
5. Compute accumulation, not just grade. Use grade multiplied by reef thickness (cm·g/t). Thin, rich and thick, lean intersections are not comparable on grade alone.
6. Grid and contour the accumulation. Use QGIS (free) to interpolate the values and look for linear, channel-shaped highs. One study notes gold and uranium are confined to channel facies and tend to accumulate in basal horizons.
7. Overlay structure. Faults displace and compartmentalise reefs. In the Witwatersrand, faults normal to the basin margins displaced the margins and form many of the boundaries of individual gold fields. Gravity and magnetics help you trace these under cover.
8. Rank and test. Drill the highest-confidence sedimentary trend first, then use the results to revise the palaeo-flow model.

Worked example (illustrative, hypothetical numbers)

Suppose a generic Archaean basin has a quartz-pebble conglomerate at an unconformity. Four historic boreholes intersect it:

• Hole A: 40 cm at 12 g/t = 480 cm·g/t
• Hole B: 25 cm at 6 g/t = 150 cm·g/t
• Hole C: 60 cm at 20 g/t = 1,200 cm·g/t
• Hole D: 10 cm at 3 g/t = 30 cm·g/t

Plotted, C and A lie on a north-east trend about 1.5 km long, and B and D sit off it. Logs show median pebble size falling from 40 mm at C to 15 mm at D, consistent with flow away from the basin margin to the north-east. The interpretation is a channel that enters near C and loses energy downstream. The next holes would step out along the trend, on the assumption that the axis continues. They would also test upstream of C toward the margin, where a thicker, coarser reef is plausible. Holes B and D would be kept as controls for the off-channel background.

This is a reasoning exercise, not a prediction. With four holes the trend is a hypothesis, and each new intersection should test it.

Common mistakes and limitations

• Treating every conglomerate as prospective. No significant gold-bearing reefs were found in the West Rand and Dominion groups. Gold needs a source, a transport path and a trap in the right sequence.
• Assuming a nugget effect does not exist. Coarse gold gives erratic assays, and a single hole can mislead. Use enough sample mass and check your repeatability.
• Ignoring post-depositional change. Metamorphism, faulting and remobilisation can shift gold locally. Do not rely on a purely sedimentary model if textures show overprinting.
• Extrapolating too far. Channels pinch out, are cut by younger scours, and are offset by faults.
• Depth. Remote sensing sees only the surface. Witwatersrand mining in buried portions is sometimes carried out at depths of 4 kilometres, so geophysics and drilling carry most of the load.

How to check your results. Test the model against holes you withheld. Confirm the palaeo-flow direction using two independent indicators, such as cross-bedding and pebble size. Verify that high values line up with a mapped surface, not a sampling artefact. Finally, check whether your trend survives removal of the single richest hole.

Key points to remember

• Sediment-hosted gold is controlled by hydraulics, source and preservation, so map the sedimentary architecture first.
• Unconformities, coarse gravel and channel facies are your primary targeting guides.
• Use accumulation (grade × thickness), not grade alone.
• Later structure and metamorphism rearrange and offset the system, so always add structural control.
• Free data (SRTM, regional geophysics, published logs) and QGIS are enough to build a first-pass model, but only drilling confirms it.

Sources

• 140 Years of Mining the Witwatersrand Basin (srk.com)
• Witwatersrand (en.wikipedia.org)
• Gold mineralization within the Witwatersrand Basin, South Africa: ... (researchgate.net)
• (PDF) A Review of the Witwatersrand Basin - The World’s Greatest ... (researchgate.net)
• (PDF) A Review of the Witwatersrand Basin - The World’s Greatest ... (academia.edu)
• (PDF) The witwatersrand gold deposit, South Africa:: An ... (researchgate.net)
• Witwatersrand Gold Deposits - 911Metallurgist (911metallurgist.com)
• Factors responsible for Witwatersrand gold mineralization A. Hofmann (ujcontent.uj.ac.za)

About Orex — Orex is a mineral exploration intelligence platform based in Mwanza, Tanzania. We combine satellite remote sensing, elevation-derived structural analysis and open geoscience data to help explorers, licence holders and investors focus their fieldwork on the ground that matters.

Want to see fault structures and intersection targets on your area of interest — for free? Install GoldRadar Faults on your phone or desktop: it maps lineaments and automatically flags fault intersections derived from satellite elevation data, giving you a structural framework for preliminary exploration before you spend a dollar on the ground.