Original publication in

Economic Geology

Vol. 69, 1974, pp. 59‑79

 

 

 

Depositional and Diagenetic Environment Related to Sulfide Mineralization, Mufulira, Zambia

 

J. G. VAN EDEN

 

Abstract

 

The Mufulira copper deposits are in the Lower Roan, lowest subdivision of the Katanga Sequence, overlying an early Precambrian Basement Complex. The Lower Roan is described as a transgressive marine sequence; at Mufulira three essentially strata bound ore bodies are contained in identical phases of a cyclic succession. Large-scale cross‑bedded units in the Footwall formation are interpreted as subaqueous dunes in a high‑energy shallow marine environment, refuting their generally accepted eolian origin. Analysis of cross‑bedding in the lower part of the Ore formation and in the Footwall formation throughout the mine reveals a bimodal distribution of paleocurrent directions with southerly offshore and northeasterly longshore directions. In the west, southerly directions with little variation around their mean vector indicate fluviatile channels.

The typical host rock of the economic mineralization at Mufulira is a fine‑grained carbonaceous wacke. This wacke was deposited in basin‑like depressions of the near shore seafloor, sheltered from the open sea by shoals. Stagnating bottom waters, resulting in anaerobic conditions, may have contributed to the exceptional conditions under which synsedimentary sulfides could form. These original sulfide deposits were probably of a low grade and were later enriched, possibly during diagenesis as is indicated by diagenetic mottling in sandstones, by mineral zoning, and by copper sulfide distribution in and around shale fragments enclosed in coarse sandstone. Clean arenites and con­glomeratie sandstones in the lowest parts of the ore bodies and in their footwall beds were ideal channelways for laterally migrating copper‑bearing fluids. The economic concentration of copper was principally dependent upon the presence of a reducing carbonaceous‑sulfureous environment that was associated with the carbonaceous wacke and a mottled sandstone.

A unique relationship of copper sulfides to a particular sedimentary facies is not observed, but sedimentological trends are consistent with the distribution pattern of the sulfides and can be used in exploration for this type of deposit.

 

 

Part 1 The Sedimentary Geology of the Footwall and Ore Formations

 

Introduction

 

THE Mufulira ore deposit is one of the major copper sulfide occurrences in the world and is a typical example of stratiform base metal deposit in arenaceous sedimentary rocks. The vast lateral extent of the essentially stratabound copper sulfides and the finely disseminated nature of ore minerals are principle features. The geologic setting is described in detail by Brandt et al. (1961) and will not be reviewed here.

All the major deposits of the Copperbelt occur in metasedimentary rocks, and though the mode of ore emplacement is uncertain, its relation to particular stratigraphic horizons, as well as to sedimentary facies, is generally recognized. Sedimentary geology has an important bearing on the understanding of base metal deposits in the Copperbelt ; a number of authors have dwelt on the subject and attempted

to interpret their depositional environment. How­ever, little rigorous sedimentologic analysis has been carried out. The depositional environment of the host rocks to the ore bodies is controversial; the consensus of opinion is that these sediments were laid down in a nearshore marine environment, but in the past these same rock types have been ascribed to such different environments as fluviatile channel and marine turbidites.

An observed sedimentary control on emplacement of the ore may be due to a direct relationship be­tween copper and the depositional environment ­assuming a synsedimentary origin‑or to a later redistribution of ore minerals which concentrated the copper in permeable or chemically receptive rocks associated with a particular sedimentary en­vironment. A similar relationship between economic mineralization and sedimentary environment may exist in other areas of the Copperbelt and in its regional extensions along the Katangan‑Damaran orogenic belt, as pointed out by van Eden and Binda (1972).

 

 Page 59 of the original publication

FIG. 1. Generalized stratigraphic, columns and graph showing trans­gressive‑regressive sedimentary cycles in the Ore formation.

 

 

 

This paper is a study of the depositional environ­ment of the Footwall beds and Ore formation based on field observations of sedimentary structures, paleocurrent analysis, and microscopic study of the sedimentary lithology. Ore genesis is discussed in the second part of the paper, using sedimentary environment as a starting point. Particular attention is paid to synsedimentary and early diagenetic ore­forming processes related to the depositional environ­ment.

 

 

Stratigraphy and Sub‑Katanga Topography

 

The ore bodies at Mufulira occur in metasedimen­tary rocks of the Lower Roan Subgroup, which is the lowest part of the Katanga Sequence (nomen­clature: Binda and van Eden, 1972). The Katanga Sequence is of Late Precambrian age and lies un­conformably on an early Precambrian Basement Complex composed of schist and granite. The un­conformity is mostly visible as a sharp contact, with large boulders of Basement rocks in the overlying metasediments. These basal conglomerates form the start of a thick transgressive series totaling many thousands of feet; the Lower Roan, Upper Roan, and Mwashia are composed predominantly of coarse clas­tics and argillites, respectively, a series which reflects increasingly deep‑water conditions (Fig. 1) .

The Lower Roan is divided into Footwall, Ore,

and Hangingwall formations, a division based on the presence or absence of copper minerals rather than on the sediment type. The sequence is cyclic, a repeated slow marine transgression being followed by a sudden minor regression. The first well developed cycle starts at the base of the Upper Foot­wall, comprises the C Orebody and ends in the Inter B/C (Fig. 1). The lithologic sequence shows a reduction in grainsize from conglomeratic sand­stone at the base, to argillitic sandstone in the ore horizon, and to dolomite and argillite at the top, a sequence which is twice repeated‑each cycle comprising an ore‑hearing horizon.

Since the sequence is repeated, it is necessary to describe one cycle only and emphasis will be laid on the Footwall formation and C Orebody, the largest of three ore bodies at Mufulira, with a strike length of more than three miles. The Upper Footwall for­mation is well exposed in the mine workings and provides data for the interpretation of the deposi­tional environment and paleogeography of the min­eralized C horizon.

Using the Mudseam, which is a thin silty dolomite at the top of the C Orebody, as datum, a plan was prepared by unrolling along dip sections from an arbitrary central Tine‑the 1,400 ft (450 m) level. The unrolling procedure introduces little distortion, since folding at Mufulira is simple with the main fold axes parallel and horizontal. The unrolled plan presents a base on which to plot paleogeographic data in close spatial relation to their original position during time of deposition.

Page 60 of the original publication

FIG. 2. Unrolled map with isopachs of Footwall plus C Orebody in feet showing Basement topography; Basement highs are reflected by low isopach values (shaded areas). Outcrop is near lower edge of diagram.

 

Isopachs of the Footwall formation and C Ore­body plotted on the unrolled plan (Fig. 2) show that the Basement topography consisted of rounded hills, ridges, and valleys, with a relief of about 400 feet. Prominent features are ridges at approximately 59 and 34 mining blocks dividing the mining area into eastern, central, and western parts. Troughs aligned in a north‑south direction at mining :blocks 72 and 43 suggest river valleys.

The Basement topography influences the paleo­current pattern, the lithofacies of the Katanga sedi­ments, and also the mineralization.

 

 

Lithology of Footwall and C Orebody Formations

 

Petrography

 

Footwall sandstones are composed of quartz and feldspar, with occasional rock fragments and ac­cessory minerals in a variable amount of fine detrital matrix. Coarse detrital feldspar is invariably micro­cline, while the matrix may contain appreciable amounts of albite or oligoclase, which is probably authigenic. Rock fragments are quartz‑sericite ag­gregaten with a fine granular mosaic structure. Most of the matrix is of primary origin and is composed of a fine granular mixture of quartz, feldspar, mus­covite, and sericite, though in part it is the product of secondary alteration such as sericitization of feld­spar and the solution of finer quartz particles fol­lowed by chemical precipitation of silica. Because of the amount of primary matrix, most Footwall sandstones are wackes and, because of their high feldspar content, the rocks are either feldspathic wackes or arkoses. Some of the cleaner sandstones have a carbonate or anhydrite cement. Unlike the

bedded anhydrite occurrences in the Upper Roan dolomites, the Lower Roan anhydrite is interstitial or forms concretionary concentrations and veins. The anhydrite may constitute a chemically precipi­tated cement that formed in the rock shortly after deposition, hut there is no indication of an evaporite environment.

Footwall metasediments of the western area are coarse, gray, and pink arkoses, feldspathic wackes, conglomerates, pink siltstones, and subordinate dark colored siltstones and argillites. The coarseness of the clastics and their feldspathic character are typical of continental clastic wedges. The more uniform feldspathic arenites and wackes of the eastern area are characteristic of peripheral marine deposits.

The C Orebody sandstones have a higher quartz content and slightly more abundant dolomitic or anhydritic cement than the Footwall sandstones. The lowest part of the C Orebody consists of felds­pathic arenites showing a more extensive recrystal­lization of quartz and feldspar; the lesser amount of primary matrix may have favored recrystallization.

The characteristic rock type of the middle and upper C Orebody is a sericitic quartzitic sandstone that is composed of detrital quartz (up to 70 percent) and feldspar in a groundmass of sericite and some carbonate cement. Referring to its primary composi­tion, this rock will be called argillitic sandstone or wacke. The normai argillitic sandstone is laterally transitional into a wacke that may contain more than one percent of organic carbon and that han a characteristic dark color. This carbonaceous wacke has wrongly been named "Graywacke."

The C Orebody sandstones have a somewhat dif­ferent composition from the Footwall sandstones, which may be due to a different provenance. The higher quartz and lower feldspar contents, however, point rather to selective weathering or attrition of the detritus, perhaps in high‑energy littoral environ­ments, before it was laid down.

 Page 61 of the original publication

 

FIG. 3 A. Photomicrograph of well‑preserved detrital grains in feldspatic sandstone (Lower Footwall) ; abundant interstitial sericite suggests an originally high clay content; large patches of sericite are altered feldspar clasts; sample 16; crossed nicols.

 

B. Photomicrograph of carbonaceous wacke; quartz clasts in matrix of copper sulfides and sericite­carbonaceous material; sample 24; plane light.

 

 Lateral variation in composition, in particular the proportion of fine matrix, reflects energy conditions in the environment. It is not clear whether the more abundant dolomitic cement as compared to the Footwall beds reflects conditions of the depositional environment or is merely a result of its stratigraphic position beneath dolomitic rocks.

 

Texture

The sedimentary rocks at Mufulira have been sub­jected to deep burial and folding and show a variable degree of recrystallization. In thin section, however, most rocks show the original boundaries of detrital quartz grains, which are distinguishable by .their "dust" rings, visible in ordinary transmitted light. Remarkably well‑preserved sedimentary textures can be observed in some specimens (Fig. 3A), hut textural analysis of the whole rock remains elusive because of the sericitization of the matrix and of clastic feldspar. Locally, cleaner sandstones have lost their original sedimentary texture almost com­pletely by recrystallization of quartz and feldspar to a mosaic texture. Quantitative measurements of textural parameters for these rocks are inevitably sub j ective.

Surface texture or even rounding of quartz grains is difficult to observe, because of pitting of the grain boundary by pressure solution and the development of overgrowths. The observed roundness varies from subangular to subrounded, without appreciable dif­ferences between the Footwall beds and C Orebody ; rounded or well‑rounded grains seem to be absent.  

 

 Page 62 of the original publication

 

FIG. 4. Size frequency distribution of long axes of quartz grains, measured in thin section; Lower Footwall (16), Upper Footwall (17, 18, 19), C Orebody lowest (23), C Orebody carbonaceous wacke (24) ; note the similarity of all Footwall sandstones, the relatively good sorting of the lowermost C Orebody and the poor sorting of C Orebody wacke.

 

TABLE 1. Grainsize Parameters of Sediments from the Mudseam, C Orebody, and Footwall Formations (Long axis of quartz grains, measured in thin section)

 

 

Grainsize measurements were made on a small number of selected samples; the data presented in Table1 are not statistically reliable, but illustrate general characteristics of typical lithologic units.

Grainsize parameters: Grainsize was measured in thin section using techniques described by Friedman (1958) . Only detrital quartz grains larger than 10 microns were measured; the exclusive use of quartz grains avoids bias from the variable amount of feld­spar that is destroyed by sericitization. Data are expressed in phi units, providing a convenient logarithmic scale that is commonly used in sedimentology. (Phi to micron conversions can be found along the horizontal axis in Fig. 4.) Grainsize parameters presented in Table I are: phi arithmetic mean (Xф), phi arithmetic standard deviation (Gф), phi skewness (SKф), and phi kurtosis (Kф) (Griffiths, 1967). All parameters are derived for the population of quartz grains computed to 100 percent, excluding the fraction smaller than 10 microns.

Mean grainsize is in agreement with the field observation ,that sandstones in the Upper Footwall beds are slightly coarser than those of the Lower Footwall beds and the middle part of the C Orebody. Though the average standard deviation is higher in the Lower Footwall beds than in the Upper Footwall beds, ‑suggesting a lesser degree of sorting for the former, differences are insignificant. The C Ore­body sediments, particularly the carbonaceous wacke, have a relatively high standard deviation, indicating poor sorting. An exception to this is the base of the C Orebody ~which like the uppermost Footwall beds is characterized by a low standard deviation ; these sandstones are rather coarse, well‑sorted arenites.

Though the significance of skewness and kurtosis questionable, these parameters clearly differentiate the sandstones. Most obvious is the low skewness and kurtosis of the uppermost Footwall and lowest C samples and the particularly high kurtosis for the C carbonaceous wacke.

Cumulative size frequency graphs for part of the data discussed above are shown in Figure 4. The graph of the C carbonaceous wacke (24) is dis­tinctly different from those of the Footwall sand­stones (16, 17, 18, 19), reflecting its poor sorting. The lowest part of the C Orebody (23) has the graph with the steepest slope, indicating relatively good sorting.

Metasediments of the Footwall beds are composed largely of sandstones with a uniform composition and texture, but subordinate conglomerates and argillite beds are important constituents. Grainsize of the metasediments has considerable lateral varia­tion, with the conglomerates abundant in the west, gradually decreasing toward the east (Fig. 5) , and thin argillite intercalations in the top of the Footwall formation more abundant in the western area. Relevant lithologic criteria may be summarized in the following idealized sequence from basal unconformity to Mudseam in the east : (1) several hundred feet of uniform Footwall sandstones, wackes, or not‑so­clean arenites, incorporating some conglomeratic lenses, (2) a few tens of feet of clean arenites in the uppermost Footwall or lowest C, (3) argillaceous sandstones or carbonaceous wackes approximately 40 feet thick, and (4) thin dolomitic siltstone, sand­stone, and argillite layers on top.

 Page 63 of the original publication

 

Fig. 5. Unrolled map showing abundance of conglomerates and conglomeratic sandstones in the Upper Footwall ; 150 feet isopach, traced from Figure 2, shows Basement highs; dashed lines show position of major mining levels and surface outcrop of Mudseam.

 

 

Sedimentary Structures

 

Types of cross‑bedding

 

Many sedimentary structures have been reported from Mufulira; Garlick (1967) describes cross­bedding, mud cracks, shale slabs, ripple marks, load casts, slumps, wash‑outs, and other (but less easily explained) structures. Cross‑bedding is the most common sedimentary structure and is easily observed throughout the Footwall beds and in .parts of the C Orebody. The foresets are often enhanced by a dis­tinctive color banding due to calcification of the coarser layers and silicification of the finer grained layers (Fig. 6). Two principal types of cross­bedding can be distinguished by their shape and scale.

Small‑scale cross‑bedding, typical of the Upper Footwall formation in the west is trough‑shaped with an average set‑thickness of 6 inches (15 cm), and numerous lenticular sets form large cosets. This "festoon" type cross‑bedding is generally considered to have formed in a fluviatile environment.

The large‑scale cross‑bedding is planar, the sets being essentially tabular. Foresets are tangential, with a long toe gently inflecting toward the base of the set (Fig. 6) . Truncation planes are remarkably straight and some units persist over several hundred feet. The set thickness of the majority of these cross‑bedded units is two to three feet, but thick­nesses up to six feet have been observed. Some of the units are grouped in cosets of two or three sets, but more than half occur as isolated units, lying between horizontally bedded sandstone. This type of cross‑bedding may be attributed to dunes in a marine shelf, tidal channel, or other high‑energy marine environment. This interpretation of environment con­trasts with the accepted eolian origin of the large­scale cross‑bedded units at Mufulira (Brandt et al., 1961; Brandt, 1962; Garlick, 1965, 1967; Fleischer, 1967) ; only Dott (1967) has expressed doubt about their eolian deposition.

The so‑called eolian sandstones form a large part of the Footwall and have greatly influenced the inter­pretation of the paleogeography and the depositional environment during Footwall and C Orebody times. Brandt (1962) assumes deposition of copper sulfides in isolated basins in a desert. Garlick (1967) shows that on the 1,150 level in the east, the C Orebody interdigitates with so‑called eolian sandstones, which are generally barren. In the light of the controversy about the origin of "eolian" sandstones and their importance to the paleogeography, it seems appropriate to discuss the large‑scale cross‑bedded units in more detail.

 

Origin of large‑scale cross‑bedding

 

In the past, the large scale of particular cross­bedded units has been taken as a criterion for eolian origin. Large‑scale cross‑bedding, with a set thickness between one and six feet is indeed most common in eolian dunes (MCBride and Hayes, 1962; McKee, 1966) ; however, large‑scale cross‑bedding is also described from modern subaqueous environ­ments. Sand waves of up to 30 feet in amplitude have been recorded for the lower course of the Mississippi River (Potter and Pettijohn, 1963). 

Page 64 of the original publication

 

 

FIG. 6. Large‑scale cross‑bedding with distinct color alterations along foresets ; Upper Footwall, eastern area.

  

 Large‑scale cross‑bédding, in units up to 40 meters thick, is allo found in deposits on shallow continental shelves (Houboult, 1968) . It is bevond doubt that cross‑bedding of equal or larger scale than that found in the Footwall beds at Mufulira can be formed in a subaqueous environment.

Steeply dipping foresets that have occasionally been observed and interpreted as eolian are also con­sidered no reliable criterion. It is difficult to dif­ferentiate on the basis of this criterion even in modern deposits (McKee, 1957). In folded rocks one may correct for the effects of tectonic tilt and axial plunge (Ramsay, 1961) and obtain accurate current directions, hut the internal deformation of beds and its influence on the inclination of foresets is difficult to assess. At Mufulira, very high angles of 40° and more in apparently undeformed rocks show the considerable steepening of primary angles. Primary angles of foresets, measured in 152 beds, average around 20°, while maximum angles can be as high as 32°. Angles in the so-called eolian beds, however, were not found to be significantly different from those measured in accepted aqueous units.

Truncation planes of the large‑scale cross‑bedded units at Mufulira are remarkably straight and mostly parallel to the normal bedding. Examples from the literature show that truncation planes in recent eolian dunes are almost always inclined (McKee, 1940). McKee (1966) finds many bounding sur­faces of sets dipping from 20° to 28° in a down­wind direction. Straight, near‑horizontal truncation planes in eolian deposits point to exceptional conditions. Stokes (1968) has explained such planes in the Navajo and Coconino sandstones as a result of sediment removal by strong winds down to the wa­ter table. His interpretation, however, refers to erosional planes that truncate composite sets of intri­cate eolian cross‑bedding quite different from the individual cross‑bedded units at Mufulira. Straight, near‑horizontal erosional and depositional surfaces are more easily explained in an aqueous environment (McKee, 1940) .

Exposed faces of Recent eolian deposits typically show numerous sets of medium‑ to large‑scale, often interlocking units of wedge or polyhedral shape (MeKee, 1940, 1966; Shrock, 1948). Many of the large‑scale cross‑bedded units at Mufulira, however, are single units sandwiched between horizontally bedded sandstones, indicating subaqueous rather than eolian deposition. In the aqueous environment the horizontally bedded sandstones originate either from sheet flow under conditions of strong currents or in a lower energy flow regime from very low amplitude sand waves (Smith, 1971) .

 

Page 65 of the original publication

 

 

The alternation of large‑scale cross‑bedded sand­stone with thin beds of conglomeratic sandstone or argillite in the Upper Footwall has been thought to originate from alternating aqueous and eolian en­vironment (Garlick, 1967) . However, without cri­teria for an eolian origin of the large‑scale cross­bedding, the definitive aqueous characteristics of some of the beds would rather point to a subaqueous origin for the whole sequence.

Lithologic characteristics may sometimes also provide reliable criteria for the recognition of eolian deposits. Eolian sand of Recent deposits is well sorted and the grafins usually show good rounding (Shep­ard and Young, 1961 ; Bigarella, 1969). Desert sandstones characteristically have well‑sorted laminae a few mm thick and commonly sharp differences in maximum grainsize between laminae (Glennie, 1970). Silt and clay fractions in modern dune sand of desert as well as coastal environments are gen­erally less than ene percent (McKee, 1966; Shepard and Young, 1961; Friedman, 1961) . The Footwall formation at Mufulira consists of poorly sorted felds­pathic sandstones with a silt and clay fraction of up to 15 percent. Well‑rounded grains and laminae of contrasting grainsize are seldom observed. Only some sandstones at the very top of the Footwall beds are better sorted and contain less matrix, hut here .the environment is generally accepted as aqueous. Though the sedimentary texture is ob­scured in the partially altered and recrystallized rocks, recognizable textural parameters exclude an eolian origin for most Footwall rocks at Mufulira.

Reinterpretation of the large‑scale cross‑bedding as subaqueous considerably simplifies the interpreta­tion of the sedimentary environment and the paleo­geography. Marine Footwall sandstones in the cast are overlain by similar C Orebody sandstones and no major change in enviroment from continental to marine has to be assumed within this uniform sequence. The amount of conglomerate in the Upper Footwall beds in the western alluvial‑littoral environment decreases towards the cast and is primarily a function of distance from the coastline in a near­shore marine environment, and absence of con­glomerates in some parts of the eastern area does not indicate an eolian origin for those sandstones. Large‑scale cross‑bedding reported by Garlick (1967) to flank the C Orebody in the south and east marks a transition to a more open, higher energy, marine environment. The implications of

this interpretation with regard to ore genesis will be discussed later.

Other sedimentary structures

Many of the smaller sedimentary structures are more difficult to recognize, but are remarkably well preserved, considering that these Precambrian rocks are folded and partially recrystallized. Ripple marks, mud cracks, load casts, etc., point to a particular depositional environment. Mud cracks, occurring in places near the base of the C Orebody in the western and central areas, show that mudflats were occasionally exposed to air, dessicated, and cracked. Shale slabs, several inches long, that are found within conglomeratic sandstone at approximately the same horizon, must have been ripped from partially indurated shale beds and laid down after only short transport.

These and other structures are characteristic of tidal mudflats in a littoral environment, they are localized near the base of the C Orebody and mark the transition from continental to marine deposits in the western area. The transition is characteristic of paralic conditions, but their subordinate occurrence between thick alluvial fan conglomerates and marine sandstones suggests that the coast must have been topographically steep and the transition of the con­tinent to the sea rapid, not allowing the development of an extensive deltaic complex.

 

 

Paleocurrent Directions

 

Paleocurrent directions and direction of sediment transport are vital factors for the analysis of the paleogeography, and directional characteristics facili­tate extrapolation of geologie data to areas as yet unexplored. Though directional structures like current ripples, scour‑and‑fill, and preconsolidation slumping occur, directional data have been inter­preted from cross‑bedding only. The correction of the direetional data for tectonic tilt has been made by a single rotation of the bedding plane about fits strike, a simple method that is acceptable since fold axes in the mining area are horizontal or plunge at less than 15°.

Paleocurrent directions in the Footwall beds fall into two distinctive groups, coinciding with the Upper and Lower Footwall stratigraphic units. In the Lower Footwall beds paleocurrents have a dominant northerly direction, but on the deeper mine levels, particularly in the west, southerly directions have been found (Hodgson, 1969). The small number of observations (approximately 100) on only two mining levels does not allow a detailed analysis, but northerly current directions indicate a net transport of sand toward the coast. A shallow marine environment is envisaged in which the cross‑bedding was formed in submarine sandbars moving land­ward similar to those described from recent shelf environments (Tanner, 1955 ; Werner, 1963) .

 Page 66 of the original publication

 

FIG. 7. Unrolled map of Upper Footwall beds showing major sedimentary environments and current directions.; dashed line indicates trend of shore line; 150 feet isopach, traced from Figure 2, shows Basement highs.

  

The Upper Footwall beds and C Orebody have consistent paleocurrent directions to the south and, less prominently, to the northeast and are therefore considered as one group. It should be stressed, however, that cross‑bedding is scarce in the C Ore­body and analysis of the paleocurrent pattern relies heavily on data of the Upper Footwall beds. The development of the paleogeography of the C Ore­body is thought to reflect gradual transition from conditions similar to those under which the Footwall beds had been deposited.

A relationship has been observed between current directions and the topography of the Basement. In the west currents are generally toward the south, following the downslope direction of a valley. Current directions over the eastern flank of the ridge on mining block 59 (Fig. 2) reflect the local topography of this hill ; current directions north of the ridge are more often to the north, while south of the ridge a dominant southerly direction is found, corresponding in both instances to a downslope direction. The ridge on block 35 and the area farther east show similar relations. The strong influence of the underlying Basement topography on current directions illustrates its controlling influence on the depositional environment of Upper Footwall beds and C Orebody. Grouped data of current directions plotted in Figure 7 show two types of distribution : (1) unimodal distributions with low variability around a mean direction toward the south, and (2) bimodal distributions with more variable directions toward

the south and a second mode to the east. The unimodal paleocurrent distributions occur mainly in the west where most cross‑bedding is small scale and forms large cosets; its fluviatile origin is unmistak­able. The great vector strength of the southerly current suggests a steep‑gradient coastal area with a sediment supply from the north. The bimodal paleocurrent directions are in the south‑central and throughout the eastern area where cross‑bedding is generally of the large‑scale planar type.

Bimodality and high variability of the paleocurrent directions are indicative of marine nearshore areas, though larger inland lakes may develop similar bi­modal characteristics. The southerly mode, which is most consistent, is parallel to the unimodal direc­tions in the western fluviatile deposits and is inter­preted as an offshore direction. The northeasterly mode is more variable. Directions of the two modes are nearly opposite at some locations in the east, indicating tidal currents. In the west‑central area current directions are found at right angles to each other, which suggests a stronger shoreline influence ; the easterly mode reflects longshore currents, while the southerly mode is the normai offshore direction. The trend of the shoreline, being at right angles to the offshore current directions, may be delineated accurately (Fig. 7), and its location was to the north of the Mufulira ore deposit.

 

 

Depositional Environment and Paleogeography

 

Footwall beds

 

The sediments of the Katanga Sequence were deposited during a major marine transgression over the early Precambrian Basement, which was the source of the Katanga sediments and must have formed an extensive land mass in the north.

 Page 67 of the original publication

 

FIG. 8. Elements of the paleogeography during C Orebody sedimentation: (1) carbonaceous wacke and (2) mottled argillitic sandstone, both (1) and (2) represent a low‑energy depositional environment in which synsedimentary or diagenetic copper sulfides could have formed; (3) and (4) coarse often cross‑bedded, sandstones of a higher energy depositional environment, littoral and off shore, respectively; 150 feet isopach, traced from Figure 2, shows underlymg Basement hills; line A‑B is location of section in Figure 9.

 

  Im­mature conglomeratic arkoses found locally at the base of the Katanga Sequence represent continental materials in situ on the ancient Basement and indicate a swift transgression. The Basement landscape was characterized by low rounded hills of schist or granite, which early in .the Katanga transgression formed small offshore islands. The Lower Footwall formation varies in thickness from 0 to 300 feet, as it fills valleys in the irregular Basement surface.

The uniform sandstone facies of the Lower Footwall beds reflects a marine environment, in which the sand was derived from older coastal or residual continental deposits. Net transport of sand, as in­dicated by the northerly directions of large‑scale crossbeds was locally toward the coast.

The Lower Footwall beds are truncated by an unconformity and overlain by conglomeratic sandstones derived from the north. The unconformity marks a minor regression, probably caused by a gentle tectonic uplift of the northern hinterland. Coarse Upper Footwall arkoses in the west were deposited in a fluviatile‑littoral environment as alluvial fans by small steep‑gradient rivers. Toward the east alluvial conglomerates become less abundant and marine sandstone prevails. A sandy coastline exposed to strong wave action is envisaged, on which offshore and longshore currents were active. The topography was now flat with most of the Basement covered by sediments, though locally it was still exposed on the seafloor. The abundance of conglomerates in valleys

in the west and near‑absence over Basement hills suggests deposition of conglomerates on channel floors in either a fluviatile or nearshore marine en­vironment. In the east conglomerates are more abundant over the low Basement hills that probably formed relatively shallow areas on the sea floor (Fig. 5) ; such shoals would have been exposed to currents and waves that winnowed out the fine sand, leaving only the coarser components. Relatively coarse sediment is not uncommon over topographic elevations in recent marine deposits. Kofoed and Gorsline (1963) show that topography is the most important single factor controlling the distribution of sedimentary material in a nearshore marine environment.

 

C Orebody formation

 

Toward the end of Footwall time the western area was transgressed by the sea and the alluvial deposits were overlain by marine sands of the C formation. In the east, where both Footwall beds and the C formation consist of marine sandstones, the start of the C is difficult to establish. The lowest beds of the C formation are clean, rather well‑sorted sand­stones that often exhibit cross‑bedding; they reflect the high‑energy environment possibly of a beach or an offshore barrier. Except for these lowest beds, the C Orebody is composed of argillitic sandstones laid down under relatively quiet conditions. As with the Upper Footwall beds the shallower parts of the depositional area were above Basement hills, which were now buried ; the topography, however, tended to be rejuvenated by compaction of the thick unconsolidated sand sequence in the valleys.

 Page 68 of the original publication

 

Fig. 9. Idealized section of C Orebody and Footwall beds during a late phase of C Orebody sedi­mentation; Basement forms a threshold, isolating part of the coastal sea floor from the open sea; line A‑B in Figure 8 shows location of section.

  

 The eastern, central, and western areas were shallow depressions on the coastal sea floor separated from each other, and from the open sea, by shoals. The sediments deposited on these shoals were relatively coarse and often cross‑bedded, while fine‑grained sand with a high percentage of silt and clay, sometimes laminated, was laid down in the tentral, deeper parts. This fine sediment reflects a low‑energy environment, below wave base, sheltered from the open sea and is the characteristic host for a large part of the ore at Mufulira. The two main topper‑bearing rock types of the C Orebody are a carbonaceous wacke and a fine mottled sandstone, to be discussed in more detail below.

Carbonaceous wacke: The carbonaceous wacke ("Graywacke") is differentiated from the normai argillitic sandstone by its Bark color, which is caused mainly by finely disseminated carbon and the high percentage of sulfides. The fine‑grained carbonaceous wacke is typically massive, but is thinly bedded or laminated in some of its lighter colored variants near its fringes. The abundante of sericitic matrix in the carbonaceous wacke and its low feldspar content as compared to the Footwall sandstones may be partly due to sericitization of feldspar clasts, as pointed out by Darnley (1960) . Several features indicate, how­ever, that the original sediment contained an appreciable amount of clayey material. The homogen­eous composition or occasional laminated structure and the absence of current indicators such as cross­bedding, imply a low‑energy depositional environ­ment where clay would be expected. The extremely poor sorting of quartz grains, as compared to other rocks at Mufulira, suggests a mixed sediment with a large clay and silt fraction. Furthermore, the abundance of finely disseminated organic carbon could probably only have been preserved in a clay­rich sediment. Present day deposits show a sys­tematic increase of organic carbon with a decrease in grainsize, and disseminated organic remains are clearly associated with clayey sediments (Strakhov, 1969).

Though a carbonaceous wacke occurs locally in the west, it forms a major sedimentary body only in the east (Fig. 8). It was formed in a slightly deeper coastal basin whose shape was inherited from the Basement (Fig. 9). The carbonaceous wacke is flanked by coarser and cleaner sandstones located over protruding Basement hills, which acted as a kind of threshold, sheltering the toast from the open sea. Thus the carbonaceous wacke was laid down in a partially isolated basin in which bottom waters may have become stagnant, providing anaerobic con­ditions in which organic matter was only partly decomposed. In comparable modern basins carbon content increases towards the center, reaching a maximum in zones of less mobile water (Strakhov, 1969). Minor amounts of detritals would be derived from the nearby littoral zone and brought in by occasional slow‑moving currents or gentle wave action.

The carbonaceous wacke has been interpreted as a channel deposit by Darnley (1960) ; Brandt (1962) supported this interpretation and suggested more specifically a deltaic environment with a river stream­ing toward the north and west. This assumed direction of flow is contrary to paleocurrent directions measured in underlying and laterally adjacent sandstones.

 

 

Page 69 of the original publication

 

 

 Fig. 10. Mottled sandstone of lower C Orebody; western area.

 

The abundante of sericitic matrix, the absence of conglomerates and cross‑bedding, as well as the geometry of the "Graywacke" body are strongly opposed to a fluviatile channel deposit, as already pointed out by Garlick (1967), who, on the other hand, favors deposition by turbidity currents. Typical turbidite successions, however, are made up of a marked alternation of argillites and sandstones, the Jatter having sharply defined bottom surfaces and usually indistinct top surfaces (Dzulynsky and taal­ton, 1965) . Sole markings such as flute casts, groove casts, or bounce casts should occur frequently. None of these characteristics is found in the car­bonaceous wacke. Graded bedding is reported by Garlick (1967), but this by itself is not sufficient proof of a turbidite environment.

Mottled sandstone: This is the lateral equivalent of the carbonaceous wacke, and apart from the mottling, it differs from the latter mainly in its lower content of preserved organic carbon. A lateral transition occurs from even‑colored to well‑mottled rock consisting of light and dark gray diffuse patches subparallel to bedding or randomly distributed (Fig. 10). Copper minerals and organic carbon are more abundant in the dark gray parts, the ligtiter parts contain more carbonate and are sometimes more porous, probably due to solution of the carbonate and anhydrite constituents.

The mottling is a result of a chemical reorganiza­tion, which probably started in the unconsolidated sediment. From recent sediments it is known that physiochemical conditions may be patchy and that mineral transformations during diagenesis may eventually result in a mottled rock (Strakhov, 1969). The white spots in the C Orebody sandstones are carbonate concretions; the rather small percentage of carbonates in these spots is normal for concretions in sandy rocks, where the coarse clastics cannot be removed easily and carbonate is limited to the pore spaces (Strakhov, 1969).

Malan (1964) and Paltridge (1968) observed the association of mottled sandstones with a stromatolite reef that occurs stratigraphically above the C horizon and they suggest that part of the mottling may be due to algal activity. Biological activity is indeed possible and algae may be a likely source for organic carbon in these Precambrian rocks.

The mottling and apparent absence of regular bedding were interpreted by Paverd (1965) as a result of submarine slumping. Garlick (1965 ; 1967) described the mottled sandstone as a "slump breccia," associated with slump and turbidity flow mechanisms. Though slump structures are observed, these are only local phenomena, and the mottling in general is not caused by slumping. The uniform texture, but variable mineral composition, shows that the mottling is a result of chemical redistribution in an otherwise undisturbed sedimentary rock.

 

 

Part 2 The Mineralization

 

Introduction

 

The association of mineralization with particular stratigraphic horizons and the intricate relations be­tween sulfides and sedimentary structures are com­monly attributed to a syngenetic origin for the ore (Garlick, 1961, 1965, 1967, 1972; Maree, 1960; Brandt, 1962; Fleischer, 1967). During the years of discovery of the Copperbelt deposits, however, the ore bodies were considered to originate from hydrothermal solutions (Gray, 1932; and others) and some authors still favor a metasomatic or other epigenetic origin on the basis of mineralogy and chemistry of the rocks and the numerous replacement and mobilization features (Darnley, 1960). As with many such controversial issues, the truth may be somewhere in between ; emplacement of the ore by a combination of synsedimentary and postsedimentary processes should not be excluded.

The generally accepted synsedimentary process leading to ore genesis is chemical precipitation of the copper as sulfides under anaerobic conditions, possibly assisted by biological processes involving bac­teria or algae. 

Page 70 of the original publication

 Such a synsedimentary origin is quite possible for shale‑type deposits, hut is it also for sulfides in arenites? At Mufulira ore minerals, mainly chalcocite, bornite, and chalcopyrite, occur finely disseminated or in coarser blebs of replacable habit in rocks ranging from argillite to coarse conglomeratic sandstone, but the typical host rock is sandstone, and the high‑energy depositional environment of some of the host rocks cannot be considered favor­able for chemical precipitation of sulfides. The mere variety of mineralized rock types suggests that not one, but a number of ore‑forming processes were active.

  

Synsedimentary Ore‑Forming Processes

 

Detrital deposition of sulfides

 

In Recent deposits copper‑bearing minerals are sometimes found along lamination planes in fine clastic sediments, and malachite or azurite, forming incrustations on organic remains, can be transported and deposited as clastic particles (Botvinkina and Yablokov, 1964). The concentration of copper transported in this way would, however, remain ex­tremely low, as pointed out by Davidson (1962). Some of the rich mineralized conglomerates and sandstones at Mufulira are reminiscent of placer deposits, but is mechanical concentration of copper sulfides possible? Minerals like diamond, zircon, gold, etc., which are mechanically and chemically more resistant than many rock‑forming minerals, are concentrated by selective weathering, but a similar process seems impossible for the relatively soluble and soft copper sulfides. Nevertheless, detrital pyrite has been reported in ancient conglom­erates of the South African Witwatersrand in con­centrations of several percent (Hoefs et al., 1968), showing that detrital sulfides can survive.

At Mufulira, thin laminar concentrations of copper sulfides frequently occur on the foresets of cross­bedding in sandstones, and visual observation may suggest a detrital origin for the sulfides. However, chemical rock constituents, such as carbonates, also occur concentrated along foreset planes; foresets have a variable grainsize at right angles to their laminae, therefore permeability will vary considerably more across the foreset planes than along them and pore fluids flowing through the rock would tend to follow coarse‑grained foreset planes. Thus, laminar concentrations of copper sulfides could have formed from rnigrating pore fluids. Permeability control on copper mineralization was sug­gested by Hamilton (1967) for a similar relationship of copper distribution and cross‑bedding at White Pine, Michigan.

Chemical precipitation of sulfides

Chemical processes may, under certain conditions, explain the precipitation of copper from solution in sea water. Biochemical processes, whereby hydrogen sulfide is liberated by anaerobic bacteria, are capable of precipitating copper as sulfides in a sedimentary environment, a process suggested by Garlick (1961, 1965) for Copperbelt deposits. Many fèatures of the mineralization are, however, more readily explained by diagenetic processes. Early diagenetic processes are inherent to any newly formed sediment and have a decisive influence on the  mineral composition of a rock. Physico‑ or bio­chemical conditions are not strictly bounded by the depositional interface; a newly formed sediment is not a closed .system, it is in contact with interstitial, as well as overlying water, and diffusion óccurs by liquid flow or ionically (Love, 1967). Chemical processes that can precipitate and concentrate copper sulfides from sea water can also precipitate copper sulfides from pore waters, while conditions of Eh and pH favorable for precipitation of sulfides (Krumbein and Garrels, 1952) are more commonly found in the sediment than in the overlying water. A pronounced exchange of elements occurs between the bottom waters and interstitial waters, so that sulfur can be found in higher concentration in pore waters than in the overlying sea water. Eventual saturation of interstitial waters may then lead to precipitation of sulfides (Larsen and Chilingar, 1967) .

The formation of iron sulfides (predominantly pyrite) by the activity of sulfate‑reducing bacteria is well known (Love, 1964). Other heavy‑metal sulfides (such as chalcopyrite) are also shown to be formed during early diagenesis (Muller, 1955; Haussiihl and Muller, 1963). Modern bottom deposits may contain colloidal metal compounds and organic carbonaceous‑sulfureous matter, from which sulfides could be formed, possibly by bacterial action. The early stages of diagenesis, in which new min­lerals are formed, occur in recent sediments to a depth of up to 10 m (Larsen and Chilingar, 1967).

The diagenetic environment below the depositional interface of a newly formed sediment is conditioned by the overlying bottom water and the type of sedi­ment. Coarse‑grained marine sandstones with open water circulation are oxygenated to well below the depositional interface, whereas in finer grained sedi­ments reducing conditions normally prevail im­mediately below the sediment surface. Though fineness facilitates deoxygenation of a sediment, it is not an absolute requirement for the formation of sulfides during early diagenesis. The critical factor in the formation of sulfides is impermeability of the sediment to oxygen migration from the overlying water, and sorting is therefore of main importance. Sulfides may develop in sandstones if they are poorly sorted and contain enough organic material (Evans, 1965 ; Love, 1967) .

 Page 71 of the original publication

  

 

FIG. 11. Fragment of shale slab, embedded in conglomeratic sandstone; base of C Orebody, western area.

 

 

 

Thus the poorly sorted and carbon‑rich argillitic sandstones and carbonaceous wackes of the Mufulira ore bodies are suitable sediments for diagenetic for­mation, and the sedimentary environment, in which these sandy host rocks were laid down, is similar to that of the ore‑shale of other Copperbelt deposits.

 

 

Features of the Mineralization and Their Genetic Interpretation

 

Mottled sandstone

 

In the mottled sandstones the light gray carbonate­rich parts are almost devoid of copper, while the dark parts are rich in copper minerals. Strakhov (1969) shows that during the growth of carbonate concretions, an impoverishment of elements like Cu must be expected in the concretionary bodies. Thus, it is not impossible that copper sulfides were present

and homogeneously distributed in the rock at the time that diagenetic carbonate concretions formed. However, the concretionary development affected the distribution of organic matter as well. Dark parts of the mottled sandstone contain abundant organic matter, and these parts of the host rock may have been particularly receptive to later introduced copper.

Though mottling occurs occasionally in the Foot­wall rocks, it is largely characteristic of ore body sandstones ; this may be explained by the high car­bon content of the Jatter sandstones. Strakhov (1969) states that the abundance of organic matter in tbe initial sediment is a critical factor in the re­distribution of carbonates. Decomposition of buried organic material generates CO₂ which facilitates solution and migration of carbonates and which eventually leads to the formation of concretions. The mottled sandstones, laterally equivalent to the carbonaceous wackes, probably originally contained an appreciable amount or organic carbon.

 

Argillite fragrnents

 

Argillite fragments occur in the conglomeratic sandstone near the base of the C Orebody in the west. The angular argillite slabs must have been ripped from partially consolidated shale beds and re­deposited after transport over a short distance. The cores of some of these argillite slabs contain microscopic disseminated bornite, amounting to about 0.5 percent in the specimen illustrated (Fig. 11) . The lighter colored borders contain less copper (Table 2), which occurs as blebs that are visible to the naked eye.

 

TABLE 2. Chemical Composition of Argillite Slab, as Percentages
Cu Fe S A1203 Mg0 CaO Total C CO2 Organic C Na20 K20	Dark colored  core 0.49 1.0 0.15 24.7 2.29 0.20 0.15 0.1 0.13 0.13 10.8	Light colored fringe 0.30 1.0 0.11 25.5 2.32 0.51 0.02 0.1 -- 0.15 10.8
These data are obtained by wet‑chemical quantitative analysis; additional semi‑quantitative spectrographic analyses on 26 other elements did not show a significant difference between the dark colored core and the light colored fringe of the specimen.

 

Page 72 of the original publication

FIG. 12. Unrolled map showing C Orebody mineralization in percent copper; numbers 1 to 5 are locations of lithologic columns in Figure 14.

 

 

The fine‑grained copper sulfides of the core coin­cide with a higher organic carbon and a lower car­bonate content than in the rim. Assuming an original uniform copper and carbon distribution before partial consolidation, the pancity of carbon round the rim would be due to oxidation by exposure to oxidizing conditions in the porous conglomeratic sandstone in which it was deposited. These oxidizing conditions could also remove copper and precipitate carbonate.

Copper sulfides in the conglomeratie sandstone have a pronounced preference for the immediate surroundings of argillite fragments; the argillite slab of Figure 11 is also enveloped in bornite and some chalcopyrite. These high concentrations of copper sulfides along the borders of argillite frag­ments may be explained as a result of a permeability barrier and the relative low Eh and pH associated with carbonaceous shale, under which conditions copper could be arrested as sulfides from alkaline and oxygenated ground waters. This process may become active immediately after deposition of the shale slab in the sand and continue until the present. The appreciable content of organic carbon and sulfide in the core of the argillite specimen indicates its continued reducing capacity. Ontward from the argillite fragment there is first, a thin rim of re­crystallized quartz associated with chalcopyrite, then coarsely recrystallized carbonate and massive particles of bornite are found. This mineral segregation across the argillite fragment and sandstone contact shows that a physicochemical gradient of outward­increasing Eh and pH may have existed at some stage, e.g., during recrystallization.

The observations suggest that (1) copper and sulfur formed synsedimentarily in euxinic clayey sediments, though in a low concentration, and (2) copper mineral constituents were moved around at more than one stage and reemplaced in or close to sediments containing organic carbon, i.e., in rocks with a relatively high reducing capacity.

 

 

Pyritic zones over Basement hills

 

Copper mineralization in the C Orebody is continuos over approximately 15,000 feet along strike except for two low‑grade gaps around mining blocks 57 and 38 (Fig. 12). These major gaps and some smaller low‑grade areas are located over Basement hills. The C Orebody thins slightly .in these areas and a change from copper sulfides to pyrite makes them uneconomic.

The pyrite zones over the Basement hills are not easily explained. The host rock reflects a depositional environment of high current and wave energy, in which the sandstones were well oxygenated. Con­ditions were unfavorable for the development of diagenetic sulfides, and the sulfides which now fill the pore spaces must have been introduced at a later stage.

An introduction of these sulfides from above is suggested by the inverted coneshape of the pyrite lenses that hang from the pyrite‑rich top of the C Orebody, and the occurrence in some locations of a thin layer of bornite between the pyrite lens and the Basement hill. Toward the end of C Orebody sedimentation more quiescent and slightly deeper water conditions prevailed, pyrite was formed throughout the area, and its constituents may have migrated downward into the hitherto unmineralized sediments that overly the Basement hills.

 

 

 

Page 73 of the original publication

 

 

 

FIG. 13. Unrolled map of  C Orebody showing generalized zones of dominant sulfide minerals.

  

Mineral zoning

 

Ore minerals at Mufulira are pyrite, chalcopyrite, bornite, and chalcocite, and though all of these are widespread, there are distinct zones in which par­ticular minerals predominate. Mineral zoning occurs stratigraphically as well as laterally. In general the stratigraphic sequence of dominant minerals is, from bottom to top: chalcocite, bornite, chalcopyrite, and pyrite. Laterally, in the C Orebody, the most obvious features are the pyritic zones that overlie Basement hills, a chalcopyrite zone that coincides roughly with the carbonaceous wacke environment, and chalcocite zones associated with fringes and near­surface portions of the ore body (Fig. 13). In the B Orebody a large central part of the carbonaceous wacke has dominant chalcopyrite mineralization. The A carbonaceous wacke mineralization is also dominantly chalcopyrite with rich chalcocite fringes. A more detailed description of the mineral zoning and its relation to lithofacies is given by Garlick (1967).

The genetic process responsible for the mineral zoning is not clear, but some comments can be made. First, the chalcopyrite of the carbonaceous wackes in all three ore bodies and the pyrite of some small occurrences of carbonaceous wacke in the western part of the C Orebody point to a relation between these minerals and lithology. Within the series of minerals here discussed, pyrite and then chalcopyrite have the strongest affiliation with a reducing environment and their synsedimentary formation in the euxinic sediments of these areas .is not impossible. Similarly, iron‑rich sulfides in the top of the C Orebody are coincident with a host rock of a quiescent sedimentary environment.

The chalcocite is associated with fringe areas and occurs as a dominant mineral in many conglomeratic sandstones underlying the ore bodies. Figure 13

clearly demonstrates its occurrence close to the sur­face in the C Orebody and for these parts a super­gene origin is obvious. Where it occurs deeper down, supergene enrichment was facilitated by highly porous conglomeratie sandstones.

A broad zonal sequence from copper‑rich to iron­rich sulfides representing depth zones parallel to the shoreline, as described by Garlick (1961), cannot be confirmed by the present author. This study of the sedimentary environment disproves such a rela­tionship between mineral zoning and distance to the shore.

Stratigraphically the fringe between the cuprif­erous zone and the overlying barren pyritic rocks is of the greatest interest. The narrow Cu‑Fe transition at the top of the C Orebody generally occurs just below the stratigraphic marker of the Mudseam, hut rises above it in the western area, as shown in Figure 14 (2). The configuration of this mineral zoning and its relation to the lithologic facies is remarkably similar to that described by White and Wright (1966) and Brown (1971) for the White Pine copper deposit of Michigan. They conclude that copper was deposited in a sulfur‑rich reducing environment, largely by replacement of original pyrite along a laterally and upward ad­vancing mineralization front. Application of such a mechanism to Mufulira would explain better the pyrite over the Basement hills as a result of insula­tion from the mineralizing solutions moving pri­marily upward from supply channels in the Footwall sandstones, since these sandstones exist only be­tween the Basement hills (W. S. White, pers. com­mun., 1973). Thus, the assumption of a White Pine model for the Mufulira deposits is tempting, but for the moment a more detailed study of the mineral zoning is required to arrive at a well‑defined answer.

  

Page 74 of the original publication

 

FIG. 14. Lithologic columns with copper assay graphs and type of mineralization; chalcocite (Cc), bornite (Bn), chalcopyrite (Cp), and pyrite (Py) ; location of columns indicated in Figure 12.

 

 Genetic Model

 

It is evident that ore‑forming processes were not limited to one phase in the history of these rocks and explaining the rich concentrations of copper sulfides requires a combination of sedimentary and diagenetic mechanisms. Ore principally occurs in two litho­types: (1) carbonaceous wacke and argillitic sand­stone, and (2) conglomeratie, often cross‑bedded arenites, locally intercalated with argillite beds. These types of ore that occur in distinct parts of the ore bodies (see identifying numerals in Fig. 16) point to diverse genetic processes and will be discussed separately.

(1) Ore in a host rock of fine‑ to medium‑grained argillitic wacke occurs as a stratiform body within the area of economie mineralization. This type of mineralization is related to a particular paleogeo­graphic realm as well as to a particular stratigraphic sequence. Stratigraphically the mineralization occurs in fine‑grained argillitic wackes that form part of a transgressive sequence, underlain by coarse alluvial or littoral sediments and overlain by deeper water marine argillites. Occurrence of the highest grade copper mineralization in the three wacke horizons of a repetitive cyclic succession illustrates convincingly the stratigraphic‑sedimentary control on the mineralization (Fig. 15) .

With respect to host rock paleogeography, this ore is bound to sediments of the nearshore marine environment; it forms a relatively narrow belt stretching along the coast, fringed on one side by the high­energy shore environment and on the other side by submarine sand bars or deeper marine environment (Fig. 8) . In some places the ore bodies are cut off by large‑scale cross‑bedded, barren sandstones. The absence of copper sulfides in these rocks need not be due to an eolian origin as proposed by Garlick (1967) ; a high‑energy marine environment flanking an area of quiescent conditions would result in a similar cutoff of mineralization, due to a rapid lateral change in lithofacies from a favorable to an unfavor­able host rock. Mineralization occurred in the rocks of marine nearshore basins, sheltered from the open sea by shoals that were located over protruding Basement hills, thus associating mineralization with a pronounced Basement topography (Fig. 9),

An original low‑grade copper content in the carbonaceous wacke may partly be explained by syn­sedimentary processes. Figure 9 shows surface waters in connection with the open sea, while bottom waters were restricted in their circulation by the shape of the sea floor hollows. Copper ions in the sea water could have precipitated as sulfides on en­countering deoxygenated bottom waters, as proposed by Garlick (1961) and others, and conditions in these bottom sediments would certainly be favorable for the formation of early diagenetic sulfides.

If the copper content of some argillites, up to 0.5 percent, may be taken as a guide to possible syngenetic copper grades, enrichment must have .taken place. During the process of enrichment these same carbonaceous wackes formed a particularly favorable host rock, due mainly to their reducing capacity.

 

page 75 of the original publication

 

FIG. 15. Typical lithologic section of Ore formation and an assay graph for copper; eastern area.

 

(2) Part of the economic mineralization is in clean, often cross‑bedded arenites or conglomeratic sandstones. The high‑energy environment in which these rocks are laid down contrasts with the quiescent conditions necessary for the synsedimentary mineralization of type (1) and it is therefore considered as a genetically different type of ore. Type (2) mineralization is always spatially associated with type (1), and it constitutes only a minor part of the volume of the ore bodies.

As mentioned above, the type (1) mineralization in argillitic‑carbonaceous sandstones must have been enriched postsedimentary, and the underlying permeable arenites could have formed an ideal channelway for mobilized pore fluids. The greater thickness of the mineralization in the west, where abundant conglomerates underlie the C horizon (Fig. 5), and the concentration of ore minerals along small unconformities such as washouts suggest a controlling influence of permeability on emplacement of the ore. The lower boundary of the C economic mineralization in the east is in well‑sorted arenites that form a unit a few tens of feet thick in the top of the Footwall formation. This lower ore boundary often follows a particular bed over some distance, but no apparent lithofacies change is observed between the mineralized and underlying barren rocks. That only the top part of the Footwall arenite unit is mineralized shows that, though permeability must have controlled the flow of copper‑bearing fluids, other factors were of importante in precipitating the copper. Favorable physicochemical conditions, such as a reducing environment associated primarily with the carbonaceous wackes, may have extended somewhat below the actual lithologic boundary within the clean arenites : water squeezed out of the overlying wackes and argillites during compaction may have influenced the diagenetic environment, a process similar to that suggested by Hamilton (1967) for White Pine.

Copper sulfides in conglomeratie sandstones of the western area are mostly conspicuously concentrated around argillite fragments (Fig. 11) or near intercalated argillite layers. The argillite intercalations interpose permeability barriers and they may have physically retarded the flow of migrating copper­bearing pore fluids. In addition, the argillite, which contained a relatively high amount of organic carbon, must have induced a reducing environment in which topper sulfides could be arrested.

  

Conclusions

 

The sedimentary succession of the Lower Roan has commonly, in the past, been described as consisting of windblown continental sands overlain by marine sediments reflecting increasingly deep water (Brandt et al., 1961). In this paper, the eolian origin of the Footwall sandstones at Mufulira is rejected and a transgressive marine succession starting at the Basement unconformity, or locally above the basal conglomerates, is proposed. The same reasoning may well apply to similar stratigraphic horizons in other areas of the Copperbelt. For instance, at Luanshya large‑scale crossbedded sandstones, previously considered to be of eolian origin, have also been reinterpreted as subaqueous by Binda and van Eden (1968) .

A reinterpretation of large‑scale cross‑bedded sandstones is not without precedent. The copper­shale deposits at Mansfeld, Germany, in many ways analogous to the Copperbelt deposits of Zambia (Garlick, 1961, p. 147), are underlain by sandstones of the Weissliegendes which for more than 50 years had been interpreted as eolian ; these have now been shown to be of marine origin (Pryor, 1971).

 

Page 76 of the original publication

 

FIG. 16. Idealized longitudinal section through Footwall and C Orebody, showing spatial relationship of (1) stratiform ore in carbonaceous‑argillitic sandstone and (2) ore in conglomeratic and cross‑bedded arenite, so‑called Footwall mineralization.

 

 

Though the Mufulira deposits contain elements that point to synsedimentary sulfide deposition, processes of enrichment during diagenesis are also evi­dent. The economic concentration of copper is, though partly dependent upon permeability characteristics, principally controlled by the reducing carbonaceous‑sulfureous environment that was as­sociated with the carbonaceous wacke and mottled sandstone. The Mufulira ore bodies may be described as containing a nucleus of carbonaceous wacke, enveloped by mottled sandstone that was, at least originally, also rich in organic carbon, and fringed by cleaner, non‑carbonaceous arenites or conglomerates. Mineralization varies from pyrite and chalcopyrite in the carbonaceous core to chalcopyrite and bornite in the enveloping rocks, and chalcocite in the (near‑surface) fringes. The deposi­tion and preservation of carbonaceous material is limited to strictly defined sedimentary environments, which may be outlined by study of the paléoge­ography. At Mufulira the carbonaceous wackes are spatially related to a shore line and to a paleotopography that was inherited from the Basement.

Uniform, large‑scale cross‑bedded sequences, underlying and laterally flanking the C Orebody in the eastern area, formed excellent permeability sys­tems, but lacked precipitants and are therefore barren. The eastern fringe of the ore body is largely determined by the sedimentary geology; mineralization in argillitic and carbonaceous sandstones of nearshore quiescent environment ends where the whole sequence below the Mudseam consists of crossbedded arenites of a high‑energy open marine environment (previously considered eolian) that occurs to the southeast. In the western area mineralization is not related to sedimentary environment; the orebody fringe occurs arbitrarily within a laterally uniform lithofacies and was probably governed by a physicochemical gradient toward the favorable carbonaceous host rocks that occur to the east. The poor correlation of the mineralization with the sedimentary environment in this area may be partly due to the effects of supergene enrichment, a process that has affected the western part of the C Orebody to a greater depth because of the locally abundant porous conglomerates.

Techniques of sedimentary petrology have shown their value in this study of the cupriferous beds at Mufulira. Major trends in the paleogeography are consistent with the distribution of the copper sulfides and the sedimentary facies is seen to have influenced both synsedimentary deposition of sulfides and the distribution of later enrichment.

 

 

Acknowledgments

 

I am indebted to F. Mendelsohn and W. G. Gar­lick who stimulated my initial interest in the geology of the Copperbelt and who outlined the direction of this study in its early stages. F. Mendelsohn also critically read the manuscript. Thanks are due to P. L. Binda, M. E. Green, and J. L. W. Dolly for helpful criticism on drafts of this paper and valuable suggestions that came forward in our many discussions about concepts of ore genesis. W. S. White of the U. S. Geological Survey reviewed the paper and is gratefully acknowledged for his constructive comments.

 

Page 77 of the original publication

The author wishes to acknowledge the Management of Roan Consolidated Mines Ltd. for permission to publish this report.

 

 

GEOLOGIC RESEARCH DEPARTMENT
ROAN CONSOLIDATED MINES LTD.
KALULUSHI, ZAMBIA

 

Present Address [ this is the address at the time of publishing the paper in Economic Geology]:

J.G. VAN EDEN
GEOLOGIC RESEARCH DEPARTMENT
JOHANNESBURG CONSOLIDATED INVESTMENT
COMPANY, LTD.
P.O. BOX 2, RANDFONTEIN REPUBLIC OF SOUTH AFRICA

February 5; August 15, 1973

 

 

REFERENCES

 

Bigarella, J. J., 1969, Dune sediments : Characteristics, recognition and importance: Am. Assoc. Petroleum Geologists Bull., v. 53, p. 707.

Binda, ‑P. L., 1969, The top of RL7 Muliashi South, Pt. II: Mineralization and sedimentary geology: Roan Con­solidated Mines, company rept. GR23.

‑ and Eden J G. van, 1968, Sedimentary structures and palaeocurrent directions in the RL7 of the quartzite quarry, Luanshya : Roan Consolidated Mines, company rept. GR12.

‑ 1971, Sedimentology of the Great Conglomerate (Tillite) at Ndola East and Itawa: Roan Consolidated Mines, company rept. GR37.

‑ 192 Sedimentological evidence on the origin of the Precambrian Great Conglomerate (Kundelungu Tillite), Zambia: Palaéogeog‑aphy, Palaeoclimatology, Palaeoecol­ogy, v. 11, p. 152‑168.

Botvinkina, L. N., and Yablokov, V. S., 1964, Specific features of deltaic deposits in Goal‑bearing and cupriferous formations, in Deltaic and shallow marine deposits: Developments in Sedimentology, v. 1, Amsterdam, Elsevier, p. 39‑47.

Brandt, R. T., 1962, Relationship of mineralization to sedi­mentation at Mufulira, Northern Rhodesia: Inst. Mining Metallurgy Bull., v. 71, p. 459‑479.

‑ Burton, C. C. J., Maree, S. C., and Woalces, M. E., 1961 Mufulira, in Mendelsohn, F., ed., The geology of the Northern Rhodesian Copperbelt: London, McDonald, p. 411‑461.

Brown, A. C, 1971, Zoning in the White Pine copper deposit, Ontonagon County, Michigan: Econ. GEOL., v. 66, p. 543‑573.

Chilingar, G. V., 1958, Some data on diagenesis obtained from Soviet literature: a summary: Geochim. et Cosmochim. Acta, v. 13, p. 213‑217.

Darnley, A. G., 1960, ~Petrology of some Rhodesian Copper­belt orebodies and associated rocks: Inst. Mining Metallurgy Trans., v. 69, p. 137‑173.

Davidson, C. F., 1962, The origin of some strata‑bound sul­fide ore deposits : Ecoti. GEoL., v. 57, p. 265‑274.

Dott, R. H., 1967, Sedimentology (two lectures), in A study visit of R. H. Dott, Sr. (by F. Mendelsohn) : Roan Consolidated Mines, company rept. GR6.

Dzulynski, S., and Walton, E. K., 1965, Sedimentary fea­tures of flysch and greywackes : Developments in Sedi­mentology, v. 7, Amsterdam, Elsevier, 274 p.

Eden, J. G. van, 1969, Paleocurrent analysis and remarks on lithostratigraphy of  "C" Orebody and Footwall For­mation, Mufulira: Roan Consolidated Mines, company rept. GR18.

‑ 1970, Paleocurrents and depositional environment of  "C" Orebody and Footwall Formation, Mufulira: Roan Consolidated Mines, company rept. GR28.

‑ and Binda, P. L., 1972, Scope of stratigraphic and sedimentologic analysis of the Katanga Sequence, Zambia

      Geol. Mijnbouw, v. 51 p. 321‑328.

Evans, G., 1965 Intertidal flat sediments and their en­vironments of deposition in the Wash : Geol. Soc. London Quart. Jour., v. 121 p 209‑245.

Fleischer, V. D., 1967, Relation between folding mineraliza­tion and sub‑Katanga topography at Mufulira mine, Zambia: Geol. Soc. South Africa Trans., v. 70, p. 1‑14.

Friedman, G. M., 1958, Determination of sieve‑size distribu­tion from thin‑section data for sedimentary petrological studies: Jour. Geology, v. 66, p. 394‑416.

‑ 1961 Distinction between dune, beach and river sands from their textural characteristics : Jour. Sed. Petrology, v. 31, p. 514‑529.

Garlick W. G., 1961, The syngenetic theory, in Mendelsohn, F., cd., The geology of the Northern Rhodesian Copperbelt: London, McDonald, p. 146‑165.

‑ 1965, Criteria for tbc recognition of syngenetic sedi­mentary mineral deposits and veins formed by their re­mobilization: Australasian Inst. Mining Metallurgy Proc., v. 6, p. 1393‑1418.

‑ 1967, Special features and sedimentary facies of strati­form sulphide deposits in arenites : Proc. 15th Inter‑Univ. Geol. Cong, Univ. Leicester, p. 107‑169.

‑ 1972, Sedimentary environment of Zambian copper deposition : Geol. Mijnbouw, v. 51, p. 277‑298.

Glennie, K. W., 1970, Desert sedimentary environments Amsterdam, Elsevier, 222 p.

Gray, A.: 132, The Mufulira copper deposit, Northern Rhodesia : Ecoti. GEOL., v. 27, p. 315‑343.

Griffiths, J C., 1967, Scientific method in analysis of sediments : New York, MeGraw‑Hill, 508 p.

Hamilton, S. K., 167 Copper mineralization in the upper part of the Copper Harbor Conglomerate at White Pine, Michigan : Econ. GEOL., v. 62, p 885‑904.

Haussehl S., .and Muller, G., 1963, Neue ZnS‑Polytypen (2R, 12R and 21R) in mesozoischen Sedimenten NW­Deutschlands : Beitr. Mineralogie Petrographie, v. 9, p. 28‑39.

Hodgson, W. A., 1969, Stratigraphy and copper distribution in the C Horizon and Footwall beds at Mufulira West: Roan Consolidated Mines, company rept. GR24.

Hoefs, J., Nielsen, H., and Schidlowski M., 1968 Sulfer isotope abundances in pyrite from the Witwatersrand conglomerates : Econ. GEOL., v. 63, p. 975‑977.

Houboult, J. J. 'H. C., 1968, Recent sediments in the South­ern Bight of the North Sea: Geol. Mijnbouw, v. 47, p. 245‑273.

Jordan, W. M., 1969, Enigma of Colorado Plateau eolian sandstone: Am. Assoc. Petroleum Geologists Bult, v. 53, p. 725.

Kofoed, J. W., and Gorsline, D. S., 1963, Sedimentary en­vironments in Apalachicola bay and vicinity, Florida

     Jour. Sed. Petrology, v 33, p. 205‑223.

Krumbein, W. C., and Garrels, R. M., 1952, Origin and classification of chemical sediments in terms of pH and oxidation‑reduction potentials : Jour. Geology, v. 60, p. 1‑33.

Larsen, L., and Chilingar, G. V., 1967, Diagenesis in sediments : Developments in Sedimentology, v. 8, Amsterdam, Elsevier, p. 1‑17.

Love, L. G, 1964, Early diagenetic pyrite in fine‑grained seiments and the genesis of sulphide ores, in Amstutz, G. C., ed., Sedimentology and ore genesis: Developments in Sedimentology, v. 2, Amsterdam, Elsevier, p. 11‑17.

‑ 1967, Sulphides of metals in recent sediments : Proc. 15th Inter‑Univ. Geol. Cong., Univ. Leicester, p. 31‑52.

Malan S. P., 1964, Stromatolites and other algal structures at Mufulira, Northern Rhodesia: Ecoti. GEOL., v. 59, p. 397‑415.

Maree, S. C., 1960 Lithology of the Mufulira copper de­posits: Assoc. African Geol. Surveys‑Symp. stratiform copper deposits of Africa, Copenhagen, p. 159‑172.

McKee, E. D., 1940, Three types of cross‑lamination in Paleozoic rocks of Northern Arizona: Am. Jour. Sci., v. 238, p. 811‑824.

 

 

Page 78 of the original publication