A Geological Model for the Alluvial Gold Environment
Mankind almost certainly first found gold when a yellow glint from the bottom of a stream bed attracted the attention of one of our ancestors in prehistoric Africa. Ever since, the allure of gold - its colour, improbable density, malleability and scarceness - mean it has been prized, and great efforts have been made to accumulate it. Most ancient peoples venerated and coveted gold and used it for decoration, and empires used gold as a store of value and a medium of exchange. The Egyptians are known to have used gold as early as about 5000 BC, followed by many others, including the Romans, the Incas, the Spaniards and, of course, the Anglo-Saxon invaders of North America, Africa, Australia and New Zealand.
Though gold was won from hard-rock deposits in ancient times most gold, until perhaps 1900, was won from riverbeds, and was traditionally called alluvial or placer gold. Prospecting for alluvial gold required relatively little equipment and always attracted hardy pioneers willing to forego the comforts of society in the hope of ‘getting rich quick’. The gold they found — if they were lucky — could almost instantly be exchanged for goods and services. Wherever the prospector led, the purveyors of overpriced food, equipment, hooch and brassy broads followed close behind.
Spain plundered, then mined, large amounts of gold, both alluvial and from hard-rock deposits, in the old Aztec and Inca empires in the first half of the 16th century, and Portugal followed up with discoveries in its African and Brazilian colonies. Russia mined gold in the Urals, and then Siberia was opened up for its alluvial gold deposits. In the same way, explorers of European origin worked their way anticlockwise round Australia (e.g. Ballarat, Bendigo and Kalgoorlie) in a series of gold rushes, and crossed the North American continent in a ceaseless quest for gold. The great alluvial goldfields of the Sierra Nevada Mountains of California were discovered in 1848, leading to the famous gold rush of the ‘49ers’. The fabulously rich Klondike gravels of the Yukon Territory in Canada, were found in 1896. Only a few years ago, a series of alluvial gold discoveries in the Amazon area of Brazil have been caught on camera and give us an insight into the mayhem that the quest for gold has caused throughout human history.
These are some of the most famous alluvial gold discoveries. However, alluvial gold has been found in most countries and the keen amateur can still try his hand with a gold pan in many of the formerly productive districts. In the UK, for example, gold grains can be panned at Helmsdale in Sutherland, in the Leadhills area of Lanarkshire or near Dolgelleau in Gwent. There are gold deposits in Ireland too, south of Dublin and in the Murrisk area, north of Connemara.
Some years ago, I worked in alluvial gold exploration and became fascinated by the facts and fiction of the old mining districts. However, there didn’t seem to be a geological model to provide an understanding of what was required to produce profitable accumulations of alluvial gold. Therefore I set about collecting and compiling the data from hard-to-access descriptions of those who had examined the famous old goldfields while they were still in production. Now, with the gold price above US $900 per troy ounce — one of the highest prices for 25 years — it seems a good time to look at how and why the old gold camps became the stuff of legend. In the following paragraphs we’ll have a look at the parameters of alluvial gold deposits, and then we’ll summarise the information in a diagram.
Clearly, there must be a primary bedrock source of gold, a zone of transportation, and a repository. However, the alluvial regime is a dynamic and extremely complex system subject to paleoclimatic, eustatic, tectonic and other distorting long-term influences. Even today, simple features are often not adequately understood.
Because alluvial gold is recovered by gravity techniques, it is essential that source areas contain an adequate supply of coarse gold that can be transported into and down drainages. Coarse gold, in terms of metal recoverable in alluvial mining operations, consists of particles never smaller than about 0.125 mm diameter and with sufficient thickness to weigh always more than about 0.02 mg. Such gold is normally derived from one of two sources:
Coarse primary (hypogene) gold such as that from the numerous deep mines of the Ballarat goldfields of Australia
• Coarse secondary (supergene) gold formed in the zone of weathering of primary gold deposits, probably under highly acid conditions of sulphide oxidation, or similarly acid conditions in laterite profiles.
The best-documented examples of this last mechanism are from Australia, but it has perhaps operated widely in laterite-belt countries such as Brazil and the Ivory Coast and may have made an important contribution to the placer gold budget. In addition, gold caught up in deposits formed in earlier cycles of erosion (the so-called ‘secondary collectors’), can be supplied to present drainages. However, accretion of fine gold into larger particles, actually within drainages, is not thought to be a significant source of recoverable alluvial gold. One of the main lines of evidence for this conclusion is that gold particles are observed to diminish in size downstream from source areas, as does the size of the associated gravel.
Four types of drainage
Alluvial gold deposits are formed in active drainages that can be classified into four types: gulches, creeks, rivers and gravel plains.
Gulches are the headwater drainages of alluvial gold provinces. They typically have steep gradients, varying from over 200 m/km down to about 25 m/km. Gradients of some rich gulches in the Klondike were 236 and 91m/km and at Ballarat rich gulch gradients were 166, 94, 77 and 23m/krn.
Creeks or streams often contain the first bodies of auriferous alluvium of any significant size and are characterised, in addition, by abrupt decreases in gradients as compared with the gulches which often feed them. Typically, creek gradients range from 30 to 6 m/km. Examples from the Klondike are 19 to 7.6 m/km in the Bonanza Creek area, 28 m/km at Eldorado Creek and 22 m/km on Dominion Creek. In the USA, Rock Creek in Wyoming ran at 35m/km and Clear Creek in Colorado at 30m/km.
Separation of rivers from streams may be somewhat arbitrary. However, auriferous rivers can be considered as forming by the coalescence of streams and can contain large productive gold deposits. They have gradients ranging from about 10 down to about 2 m/km. In the Klondike, the Indian River ran at 3.4 m/km and the Klondike River at 2.8 to 2.3 m/km, while in Ballarat, the paleo-Yarrowee ran at 10 m/km.
Where a river washes out into an unconfined valley or coastal plain, its carrying capacity is suddenly reduced and a gravelly plain forms. Gradients here are typically 3 to less than 1 m/km. Examples, lacking in precise data, are the American and Yuba rivers entering the Sacramento valley in California and, possibly, the River Nechi at El Bagre in Colombia.
Gulches or Gullys
Gulches receive most of their load by mass transport (slides, creep and solifluction) of colluvium from the steep slopes which border them. Gully debris is characterised by boulders (over 26.6 cm across) and blocks often up to several meters long. Since gullies are zones of active down-cutting they do not normally retain the colluvium, except locally if bedrock traps are present or if debris jams occur which provide shelter for finer material. The copious amounts of clay, often present in colluvium, are largely washed straight through the gully environment and clay is practically absent from gully deposits. Only small - though sometimes rich - gold accumulations survive and
these are invariably worked by artisenal methods.
Creeks, as already mentioned, sometimes contain significant quantities of alluvium capable of supporting bulldozer/scraper operations typically of 2,000 to 3,000 m3/day. Creek gravels are often 1.5-5 m thick and there is often a sandy or clayey overburden. Valleys can vary from a few meters to perhaps 200 m wide. Debris commonly contains boulders where there is an appropriate competent source rock, but cobbles (25.6 to 6.4 cm in diameter) are more typical. Given these characteristics, some creeks can support dredging operations.
Bonanza Creek, in the Klondike, was dredged over its lower 14 km. It was usually between 90 and 180 m wide, locally up to 275 m, the width increasing gradually, but irregularly, down the valley. The gravel layer was fairly uniform, from 1.2 to 2.4 m in thickness all across the flat bottom of the valley, and was overlain by an organic, silt-rich layer from 1.5 to 4.6 m thick. The gravel comprised clean, flat, fairly well worn pebbles, from 2.5 to 15 cm in length and 2.5 to 5cm in thickness. This was derived from the micaceous schists of the area, associated with rounded and sub-angular pebbles of quartz and, occasionally, large quartz boulders, usually angular in form.
Rock Creek, in Wyoming, ran in a valley 30 to 76 m wide and carried workable gravels over 22.5 km. A layer of gravel 2.7 to 3.7 m thick was overlain by a metre of barren loam. The gravel was well-rounded and contained only a few boulders, up to about 46 cm diameter
- 65% of the gravel passed through a screen with openings of 2 x 3.8 cm.
Various auriferous creeks were worked in Surinam. Gradients in the lower reaches were anomalously low at 2.5 to 4 m/km, because the creeks are base-levelled to the major Lawa River, of which they are tributaries. Valley floors are correspondingly wide, up to several hundred meters, and flat, and the Lawa River backs up into them in times of flood. The creeks contained reserves quoted at 0.4 to 3.2 M m3 (million cubic metres) at grades ranging from 400 to 680 mg! m3. This would be the grade from surface down to bedrock, that is the average grade of all the ground excavated in order to extract the gold. A 2 m layer of silt/clay overlay a layer of gravel 0.5 to 1 m thick comprising almost exclusively resistate angular quartz pebbles 6 to 7 cm in diameter; locally, quartz blocks of 25 cm diameter were common. Sand and clay sometimes occurred in the gravels.
At Slate Creek in Alaska, the gravels sometimes reached a thickness of 25 m.
In general, creeks can be considered to have a resource of payable alluvium ranging from 0.5 to as much as 20 M m3.
Rivers carry significantly larger alluvial resources than creeks, with finer grain sizes and presence of more clay in the overburden. Boulders are often rare except along the valley sides and/or in the case of mountainous torrential environments and/or in glaciated areas (e.g. Lonquimay in south-central Chile). Rivers often represent ideal dredging ground.
An example of a large river alluvium is that of the River Jequitinonha in Brazil, dredged for diamonds and gold. This varies from about 400 to about 2,000 m wide, locally reaching 3 km, and has been worked discontinuously over a distance of 100 km. It is fed laterally by an auriferous and diamondiferous Precambrian secondary collector and the river therefore flows within its source area. This means that equilibrium between gold (and diamond) supply and deposition may not strictly be reached. Depth to bedrock is usually from 10 to 17 m of which up to 13 m is gravel and up to 12 m is sand overburden. The ratio of sand to gravel is approximately 2:1. The gravel comprises well- rounded quartz and quartzite of poor sphericity, normally up to about 10 cm long. Gravel, larger than about 20 to 30 cm is quite rare, except near the edge of the alluvium where large scree blocks of bedrock occur.
In general, gravel plain environments have rarely been found to carry economic values as the gold tends to be dispersed and grades diluted. They are typified by unconfined alluvium, enormous volumes of sands and fine gravels and very flat gradients. In the Sacramento valley of California, the proximal parts of a gravel plain gave rise to very large, economically dredgeable, gold deposits. The Yuba dredge field extended a short distance out into the Sacramento valley and probably aggregated more than 200 M m3. The Folsom field stretched into the valley over 11 km with a width of 1.6 to 3.2 km and contained at least 430 M m3 of profitable ground.
The gold dredging operation on the Rio Nechi near El Bagre in Colombia formerly worked a very large unconfined alluvium 1.5 to 2 km wide over a distance of 20km, downstream of which values became uneconomic. Prior to dredging, the river had a meandering course with oxbow lakes. A topstratum silt and clay layer 6-15 m thick overlay 18 to 27 m of auriferous gravels, which rarely reached cobble size.
The Nature of the Gold
Gulch gold is the coarsest that exists in any part of a river system. If nuggets (pieces of gold weighing more than 0.1 g) are present, they will mostly be found in gulches (narrow ravines), provided suitable traps are present, such as irregular bedrock. In gulch alluvium, the vast majority of the gold will be found on, or in crevices within, the bedrock. Gulch gold is often coarse and angular and may contain silicate debris, especially quartz..
As examples, gold from Victoria Gulch on the Klondike was described as “sharply angular”. In the Ballarat gullies, some enormous nuggets were found and Canadian Gully yielded nuggets of 50.4, 34.7 and 31.4 kg. At White Horse Gully in Bendigo a 17.8 kg nugget (including some quartz) was found. (Interestingly, of a list of 92 Victorian nuggets, 34 came from localities specifically named “gullies”.) Finally, in the Sierra Nevada of California, most of the gold is from gulches or minor streams close to croppings.
Creek gold can also include coarse material and the payable values are, in any case, normally concentrated on bedrock, forming the so- called “autochthonous placers”. The famous Klondike creeks are a classic example of this.
For Dominion Creek (above Lombard Creek), the gold occurred as rough, rounded grains and small nuggets. Further downstream, mixtures of heavy grains were found. Some were well worn and others quite rough. There was also a more flaky variety and an occasional large, well-worn nugget. Even further downstream was more flaky gold with only occasional nuggets.
For Sulphur Creek, the gold followed the general rule in occurring as large angular pieces in the upper gulch part of the creek and in small, flaky, rough grains further down.
For Clear Creek, in Colorado, the gold was described as ranging in size from “fine” to “coarse”, but most of it was coarse.
In the Lawa River creeks of Surinam, nuggets of up to hundreds of grams were found, although certain drainages were characterised by “fine” gold. However, the gold was generally described as “coarse” and pay- streaks carrying gold values of 3 to 4 g/m3 of gravel are recorded. Gold particles of up to 347mg were recovered from exploration drill holes.
All the Klondike creeks (where the gravels were frozen with permafrost), all the Ballarat deep leads and Clear Creek, in Colorado were initially mined underground by drifting along bedrock - a sure sign of the presence of autochthonous gold. (The word “autochthonous” means “formed in the region where found”.) In the Klondike creeks and Clear Creek, reworking from surface later proved successful, partly because a proportion of the gold was above the bedrock and within the gravels (being “allochthonous” or “found in a place other than where formed”). Other creeks contain almost exclusively allochthonous gold, for example, Slate Creek in Alaska. However, there is a suspicion that Slate Creek is big enough to be classed as a “river”.
Rivers and gravel plains
Gold in rivers and gravel plains is finer than creek gold and is sometimes exclusively allochthonous and/or above-bottom autochthonous, as in the gravels of the River Nechi in Colombia. There, the best values sometimes occurred at the top of the gravels (beneath the clay top-stratum), sometimes in the middle and sometimes at the base. Storm gold is still carried down the river at times of flood and is deposited on meander and point bars, where it is worked by expert women goldpanners. The upstream gravels carried coarser, higher grade gold of lower fineness (typically containing a greater proportion of silver) than those worked further downstream. Over the 20 km stretch dredged in 1974, the amount of jig-recoverable gold dropped from 170 to 110 mg/m3 while, at the same time, the proportion of fine gold increased downstream.
The Yuba River deposits of California have been dredged several times by deeper-digging equipment, because the gold was distributed throughout the gravel column.
The examples cited above are obviously cases where all the necessary conditions for the formation of economic alluvial gold deposits were fulfilled. However, in most cases where particles of gold are present in drainages, a successful economic grade/volume combination is not present and it is interesting to consider why this may be. Probably, the most common reason is the inexistence of sufficient coarse gold in the underlying rocks. Sometimes there are important hard-rock mines in the area, such as at Ballarat or in the Sierra Nevada of California. Sometimes, there are no significant hard-rock mines at all, such as in the Klondike. However, for placer gold deposits to form, there must be an adequate supply of gravity-recoverable gold to the drainages.
The Snake River, in the northwest USA, is a famous example of a well-mineralised drainage, where a total lack of coarse gold has resulted in the failure of numerous attempts at commercial recovery. In other cases, there may be plenty of coarse gold entering the drainage system, but this is then diluted by barren material brought in by other drainages. For example, all the north- bank tributaries of the River Tagus in Portugal have been intermittently worked for gold and there are local workings on the north-bank terraces of the main river itself. However, the main river that flows for many kilometers across Spain and its south bank tributaries have very little gold and the main alluvium is uneconomic. The high probability of dilution in large drainage systems is perhaps a principal reason why comparatively few river alluviums are workable.
Gold is the least transportable component of alluvial debris, unless it occurs as small, scaly flakes, in which case it will travel far. Coarser gold is transported only in the bed load environment and, due to its small relative size, tends to get trapped in pebble or cobble- supported openwork gravels: in effect, open-work gravels work like a “jig bed” vibrated by the river when in flood.
In areas of subdued relief, especially Precambrian shield and platform environments, coarse gold cannot be transported beyond the gulch and creek environment. Allochthonous gold is transported under these conditions, but traps tend to be shallow and the gold deposits are transitory.
A good example is the River Tapajos province in the Amazon area of Brazil (e.g. the River Jamanxim and its tributary the River Novo) where innumerable gold camps exist in gulches and creeks in which gradients are quite normal and beautiful, coarse gold (including large nuggets) is plentiful. Gold is distributed throughout the sands and minor gravels of the rivers. However, bedrock concentrations appear to be absent and the alluvium available is in thin sheets with a thick clay top-stratum.
Similarly, the Lawa River in Surninarn runs on bedrock with rapids being interspersed with shallow alluvial basins.
A third example is provided by part of the Batang Hari in Sumatra. The area comprises part of a meander belt and is characterised by a top-stratum of clays and silts overlying a thick substratum of gravels and sand. Exploration drilling failed to find bottom-enrichment and the gold in the gravels was no different from that occurring on the surface.
In contrast, if the gradient is too steep, dilution is excessive, sorting of the alluvium is poor or lacking, boulders are transported far downstream and gold, even if coarse, is rarely concentrated and may be inaccessible. Such situations are common in the Andes where gradients of even major rivers can be extremely steep.
As an example, the fall of the River Lluta in northern Chile (though practically non-auriferous) ranges from 52 m/km at 70 km from the sea down to 15.4 m/km at the coast. In fact, boulders are carried all the way down to sea level. Here, concentration is further hindered by the torrential (wadi-type) climate.
In the case of the River Santa in Peru (which carries uneconomic allochthonous gold values), gradients are as much as 7.5 m/km even where the flood plain is 2 to 6 km wide.
However, alluvial gold has been successfully worked from a steep river system in Bolivia. The River Tipuani has a gradient of 32 m/km from Lambramani down to Tipuani and is characterised by ill-sorted boulder and cobble gravels. The best values are concentrated on bedrock (e.g. 8 g/m3) amongst plentiful boulders, but above-bottom concentrations and allochthonous gold also exists. The gold is coarse, flat and flaky with particles often being 4 to 6 mm in diameter. To find economic concentrations in such a boulder-strewn environment as the Tipuani is exceptional. However, at Tipuani there is a reworked paleo-channel deposit, known as the “Cangali”, which was very rich and may have formed under different physical conditions from today. Downstream of Tipuani, dredging was carried out, but conditions were extremely difficult and the ground full of boulders.
In general, higher gradients have the effect of transporting the high-energy gulch environment down into creeks and rivers.
Post depositional effects
Following formation, alluvial gold deposits may form terraces as a result of being uplifted and dissected, or they may be buried under accumulations of alluvium, tuffs, lavas, or other rock types forming the so-called “deep leads”. In practice, gulch deposits are those most likely to be eroded and gravel plain deposits the most likely to be preserved. Creek and river alluviums are those most usually found forming terrace deposits. In general, gold grades are enhanced if multiple cycles of reworking take place after deposition (e.g. the Sierra Nevada and the Otago area of New Zealand) and, in particular, by the reworking of mature landscapes in the source area which then become dissected (e.g. the Klondike). This is a very positive feature for the formation of economic placer deposits.
In the Sierra Nevada of California, on the River Duerna, and at Las Medulas in northwest Spain and in many other places, terrace deposits have been worked successfully. In fact, the Spanish deposits were the subject of vast operations by the Romans.
The semi-quantitative model for the alluvial gold environment described here is based on a number of key facts collected from areas that have provided profitable gold production in years gone by. Knowledge of the gradients, size of drainage and type of sediment, together with an understanding of the size and distribution of gold particles within productive deposits should provide greater interest and understanding when visiting auriferous areas. It does not take too much imagination to develop practical activities that can be undertaken in such areas.
For example, using a good topographic map, it is quite easy to work out the gradient of gold-bearing streams to see if they fall within the “productive” ranges. It is also possible to spot the most promising sample-locations to improve the chances of finding more and larger gold grains. This can be done perhaps by sampling the base of a terrace deposit exposed on the valley side, or by taking great care to scoop out the heavy sediment trapped on, or in cracks in, bedrock in gulleys.
The more serious practitioner might also find it helpful to have information about grades of alluvium that have paid in the past, and to know a little more about the implications for equipment of boulders in the gravels or the dilution that might be caused by barren drainages in the area. The model may not apply, or may not be directly applicable, to prospects within the source area because the necessary sedimentary equilibrium is constantly upset. This might be the case, for example, where extensive secondary collectors occur, such as those typically found in glacial outwash zones and in some sedimentary conglomerates.
What I can say is that had the information that supports this model been available to me some years ago, it would have been helpful in focusing on better prospects and, who knows, with gold prices once again making headlines, it might have a practical value in the future.
Written by Philip Dunkerley
This page was last modified on 13 May 2009
A version of this article was published in Deposits International Rock & Fossil Magazine in Issues 7 & 8,2006. ISSN 1744-9588.