Skip to Main Content

Corresponding author: e-mail, d.selley@utas.edu.au

Abstract

The Central African Copperbelt is the world’s premier sediment-hosted Cu province. It is contained in the Katangan basin, an intracratonic rift that records onset of growth at ~840 Ma and inversion at ~535 Ma. In the Copperbelt region, the basin has a crudely symmetrical form, with a central depocenter maximum containing ~11 km of strata positioned on the northern side of the border of the Democratic Republic of Congo and Zambia, and marginal condensed sequences <2 km in thickness. This fundamental extensional geometry was preserved through orogenesis, although complex configurations related to halokinesis are prevalent in central and northern parts of the basin, whereas to the south, relatively high-grade metamorphism occurred as a result of basement-involved thrusting and burial.

The largest Cu ± Co ores, both stratiform and vein-controlled, are known from the periphery of the basin and transition to U-Ni-Co and Pb-Zn-Cu ores toward the depocenter maximum. Most ore types are positioned within a ~500-m halo to former near-basin-wide salt sheets or associated halokinetic structures, the exception being that located in extreme basin marginal positions, where primary salt was not deposited. Stratiform Cu ± Co ores occur at intrasalt (Congolese-type), subsalt (Zambian-type), and salt-marginal (Kamoa-type) positions. Bulk crush-leach fluid inclusion data from the first two of these deposit types reveal a principal association with residual evaporitic brines. A likely signature of the ore fluids, the brines were generated during deposition of the basin-wide salt-sheets and occupied voluminous sub and intrasalt aquifers from ~800 Ma. Associated intense Mg ± K metasomatism was restricted to these levels, indicating that capping and enclosing salt remained impermeable for prolonged periods of the basin’s history, isolating the deep-seated aquifers from the upper part of the basin fill.

From ~765 to 740 Ma, the salt sheets in the Congolese part of the basin were halokinetically modified. Salt was withdrawn laterally to feed diapirs, ultimately leading to localized welding or breaching of the former hydrological seal. At these points, deeper-level residual brines were drawn into the intrasalt stratigraphy to interact with reducing elements and form the stratiform ores. It is probable that salt welding occurred diachronously across the northern and central parts of the basin, depending upon the interplay of original salt thickness, rates and volumes of sediment supply during accumulation of salt overburden, and tectonism. The variable timing of this fundamental change in hydrologic architecture is poorly constrained to the period of halokinetic onset to the earliest stages of orogenesis; however, the geometry of the ores and associated alteration patterns demands that mineralization preceded the characteristically complex fragmentation of the host strata. Thus, while an early orogenic timing is permissible, mineralization during the later stages of extensional basin development was more likely.

In situ reducing elements that host Zambian-type stratiform Cu ± Co ores were in continuous hydrological communication with subsalt aquifers, such that ore formation could have commenced from the ~800 Ma brine introduction event. The nonhalokinetic character of the salt in this region allowed the intact seal to have maintained suprahydrostatic pore pressures, facilitating fluid circulation until late stages of basin growth and possibly early stage orogenesis.

Leachate data from ores positioned in the depocenter maximum and southern parts of the basin that underwent relatively high grade metamorphism record mixing of residual and halite dissolution-related brines. Salt dissolution was likely triggered by emergence of diapirs or thermally and/or mechanically induced increased permeability of halite. While it is certain that halite dissolution occurred during and after orogenesis, conditions favorable for salt dissolution may have existed locally during extension in the depocenter maximum. The permeability of salt increased to a point where it became the principal aquifer. The salt’s properties as an aquiclude lost, originally deep-seated residual brine mixed with new phases of evaporite dissolution-related brine to produce ores at middle levels of the basin fill. During the final stages of ore formation, recorded by postorogenic Pb-Zn-Cu mineralization in the depocenter maximum, the salinity of fluids was dominantly derived from the dissolution of remnant bodies of salt.

Introduction

This paper presents a broad-picture view of the structural and hydrologic architecture of the world’s premier sedimenthosted stratiform Cu province. Its intention is to erect a robust macroscopic framework to account for the distribution of diverse deposit types in both time and space and in doing so provide a useful guide for exploration. Much of the data and many of the concepts presented were produced during the period of the AMIRA-ARC P544 and P872 projects (2000–2010), which focused on the Zambian and Congolese parts of the Central African Copperbelt, respectively, and built upon a wealth of existing knowledge. Since the projects’ completion, several of the authors have continued research and exploration activities throughout the Central African Copperbelt. Moreover, there have been numerous important contributions, in terms of ore deposit characterization, fluid composition and evolution, and tectonic development of the region (e.g., El Desouky et al., 2009; Bernau et al., 2013; Schmandt et al., 2013; Woodhead, 2013; Eglinger et al., 2014, 2016; Nowecki, 2014; Capistrant et al., 2015; Turlin et al., 2016).

Several recent publications on the Central African Copperbelt have reignited long-standing arguments about the timing of mineralization (Muchez et al., 2015, 2017; Sillitoe et al., 2015, 2017a, b; Hitzman and Broughton, 2017). The root of the controversy is the apparent contradiction between relative timing constraints based on textural, geometric, and compositional features of the ores and their host strata and certain absolute ages derived from the Re-Os molybdenite geochronometer. Although we also address this issue, robustly constraining the age(s) of mineralization is not our primary focus; rather we identify the stages of basin evolution during which mineralization was permissible. In this respect, it is important to appreciate that while absolute age dating underpins exploration of magmatic-hydrothermal systems, where the time-dependent fertility of magmas is paramount, it is less crucial for sediment-hosted Cu deposits, where the distribution of deposits is fundamentally controlled by chemical and permeability frameworks upon which ore fluids were superimposed.

This paper first considers the diverse structural styles of the Katangan basin, host to the Central African Copperbelt. We assess whether the fundamental geometric integrity of the basin is preserved or, as conventionally interpreted, it is now an amalgam of parautochthonous and allochthonous thrust slices (e.g., Porada and Berhorst, 2000). The analysis is based both on deposit-scale observations (our own observations and those of numerous published studies) and on published and propriety district- to regional-scale map sets, in particular the detailed 1:20,000 Gecamines series from the Democratic Republic of Congo. The subsequent section considers the form and origin of macroscopic alteration patterns within the context of the structural model. In addition, we incorporate previously unpublished Na-Cl-Br fluid inclusion data with data sets that occur in the literature (Heijlen et al., 2008; Nowecki, 2014) in an attempt to constrain the evolution of ore-related brines.

Geologic Setting

The Neoproterozoic Katangan basin forms one of a series of inverted rift basins fringing the Congo and Kalahari cratons. Collectively, they record the dispersal of Rodina from ~900 Ma and final amalgamation of central Gondwana by ~500 Ma. While the adjacent Damara and Mozambique belts contain fragments of ocean crust that are indicative of successful rifting, there is no compelling evidence that the Katangan basin ever developed beyond an intracontinental rift.

Stratigraphy and basin growth

Rocks of the Katangan basin are assigned to the Katangan Supergroup, a succession of continental, marginal marine, and subwave-base marine metasedimentary units and volumetrically minor mafic igneous strata (Cailteux et al., 1994, 2005; Bull et al., 2011). The succession is divided into three regionally mappable groups, which from oldest to youngest are named the Roan, Nguba, and Kundelungu Groups. Unconformably underlying basement strata comprise multiply reworked Archaean to Mesoproterozoic metagranites, migmatites, and metavolcanic and metasedimentary units and relatively weakly deformed Neoproterozoic metagranites. An 877 ± 11 Ma U-Pb zircon crystallization age for the youngest basement granitic suite (Armstrong et al., 1999) provides a maximum age limit for the Katangan Supergroup. Basement rocks crop out as a series of disconnected, major subbasin-fringing inliers that are conventionally grouped into four domains: anticlockwise from northern Zambia into the Democratic Republic of Congo, these are the Kafue anticline, Domes region, and Kibaran and Bangweulu blocks (Fig. 1). Copper mineralization extends beyond the mapped limits of these inliers; however, the vast majority of the Central African Copperbelt Cu resources occur either close to or inboard of outcropping basement (Fig. 1). Throughout this paper, the term Katangan basin refers to this particularly well Cuendowed basin compartment. Analysis of regional map data sets indicates that the Katangan Supergroup attains thicknesses of ~11 km within the compartment’s depocenter maximum, a region inboard of the basement inliers that occurs largely within the Democratic Republic of Congo, but as little as ~1 to 2 km at the fringes of the inliers (Figs. 1, 2).

Fig. 1.

Basic geologic map of the Central African Copperbelt, showing the distribution of various ore types. Abbreviations: CCB = Congolese Copperbelt, DRC = Democratic Republic of Congo, Gp = Group, PGE = platinum group element, Sgp = Subgroup, ZCB = Zambian Copperbelt. Inset shows map location in Africa.

Fig. 1.

Basic geologic map of the Central African Copperbelt, showing the distribution of various ore types. Abbreviations: CCB = Congolese Copperbelt, DRC = Democratic Republic of Congo, Gp = Group, PGE = platinum group element, Sgp = Subgroup, ZCB = Zambian Copperbelt. Inset shows map location in Africa.

Fig. 2.

Stratigraphic framework of the Katangan basin, highlighting thickness contrasts between the central depocenter maximum and peripheral domains where basement inliers are exposed. Generalized locations of the three profiles are shown in Figure 1. The stratigraphic framework for the Roan Group is that of Bull et al. (2011); see text for further discussion and alternative interpretations. Stratal thicknesses depicted in peripheral domains are minima for the respective areas and determined from drilling and analysis of maps. The profile of the depocenter maximum is derived from analysis of Gecamines 1:20,000 maps (noncontiguous maximum thickness measurements for individual formations are shown for middle and upper basin levels), and a gravity-constrained permissible thickness of the nonexposed synrift level of the Lower Roan Subgroup (see Digital App. Fig. A1). A conservative maximum thickness of ~11 km is determined from the balanced cross section shown in Figure 16. Abbreviation: DRC = Democratic Republic of Congo.

Fig. 2.

Stratigraphic framework of the Katangan basin, highlighting thickness contrasts between the central depocenter maximum and peripheral domains where basement inliers are exposed. Generalized locations of the three profiles are shown in Figure 1. The stratigraphic framework for the Roan Group is that of Bull et al. (2011); see text for further discussion and alternative interpretations. Stratal thicknesses depicted in peripheral domains are minima for the respective areas and determined from drilling and analysis of maps. The profile of the depocenter maximum is derived from analysis of Gecamines 1:20,000 maps (noncontiguous maximum thickness measurements for individual formations are shown for middle and upper basin levels), and a gravity-constrained permissible thickness of the nonexposed synrift level of the Lower Roan Subgroup (see Digital App. Fig. A1). A conservative maximum thickness of ~11 km is determined from the balanced cross section shown in Figure 16. Abbreviation: DRC = Democratic Republic of Congo.

Two major rift phases are evident in the stratigraphy. The first is recorded by dominantly siliciclastic strata of the Lower Roan Subgroup (Fig. 2) that is best known in Zambia. Initial terrigenous sedimentation occurred during a synrift stage of accommodation development, involving partly connected, fault-bounded depocenters (Selley et al., 2005). About the Kafue anticline, subarkosic and conglomeratic synrift strata vary in thickness from 0 to ~1,000 m (avg ~300 m). Although much more highly deformed, similarly variable thicknesses of synrift strata mantle basement inliers in the Domes region. An abrupt thickening at this level occurs at least locally immediately to the north of the inliers (~2,000 m; First Quantum Minerals, unpub. drill hole data), in the direction of the depocenter maximum.

On the western flank of the Kafue anticline, rift climax is recorded by an abrupt marine transgression and the deposition of formerly organic-rich, partly evaporitic, shale and siltstone facies of the ~25-m-thick Copperbelt Orebody Member (Fig. 2; Selley et al., 2005). Reintroduction of coarser-grained siliciclastic detritus occurred during more subdued accommodation development associated with a postrift episode, before passing into a period of basin starvation and evaporite deposition (Upper Roan Subgroup). Preserved evaporites are dominantly sub- and intertidal carbonates (anhydrite-bearing dolomites and subordinate magnesite; Woodhead, 2013), but numerous stratabound and transgressive breccia bodies have been interpreted to represent the residuum of formerly volumetrically significant halite, now withdrawn and/or dissolved (Selley et al., 2005; Hitzman et al., 2012; Woodhead, 2013).

Projection of initial rift phase elements into the depocenter maximum is hampered by nonexposure of lower stratigraphic levels (Fig. 2). Here, the basin is decoupled at an evaporitic interval deep within the structural profile (Jackson et al., 2003). Bull et al. (2011) interpret this phase of evaporite deposition to correlate with rift climax strata exposed to the south (i.e., older than Upper Roan Subgroup salt), arguing that the structural reconfiguration during this event marked the first time in the basin’s history for significant volumes of shallow marine waters to be isolated. Underlying synrift strata are never exposed, with the exception, perhaps, of an arkosic megablock plucked from subdecoupling levels in the central part of the depocenter (Derriks and Vaes, 1956). Regional gravity data permit synrift strata thicknesses to exceed 3 km within the depocenter maximum (Digital App. Fig. A1; Duffett et al., 2010). Stratigraphically overlying the lowermost evaporitic interval in the Democratic Republic of Congo is a succession of oxidizing argillaceous arenites and variably reducing inter- and subtidal carbonate and siltstone, the latter facies association forming the ~150-m-thick Mines Subgroup (Fig. 2; Cailteux et al., 1994). The original thickness of oxidizing arenites that directly underlie the Mines Subgroup is difficult to constrain due to basal decoupling; however, a maximum preserved thickness of ~300 m is known from the Kolwezi area (A.P. François, unpub. report, 1973; Fig. 1).

Correlation between the Zambian and Congolese sequences has been contentious. Whereas Mendelsohn (1961) and Fleischer et al. (1976) correlated the dolomitic Upper Roan Subgroup in Zambia with the Mines Subgroup, implying that the latter and the Copperbelt Orebody Member represent stratigraphic entities, François (1974) and Cailteux in a series of publications (Cailteux and Lefèbvre, 1975; Cailteux et al., 1994) argued on lithostratigraphic grounds that the two reducing intervals are lateral facies variants. Utilizing a combined sequence- and chemostratigraphic model, Bull et al. (2011) concluded that the Copperbelt Orebody Member and Mines Subgroup are stratigraphically distinct, correlating the latter with the upper levels of the Lower Roan Subgroup. Woodhead (2013), also using a sequence stratigraphic approach, favored correlation of the Mines Subgroup with a part of the Upper Roan Subgroup.

Despite the lack of consensus on the exact relative positions of the Copperbelt Orebody Member and Mines Subgroup, both the Bull et al. (2011) and Woodhead (2013) models agree that the latter deposited during the initial postrift stage. Moreover, stratabound breccia facies bound and punctuate the Congolese sequence (Fig. 2), implying that it was originally enclosed in salt, rather than simply capped by salt in the case of the Copperbelt Orebody Member. There is agreement between all workers that the Congolese Dipeta Subgroup, a breccia-enclosed, carbonate- and arenite-dominated interval that overlies the Mines Subgroup (Fig. 2), equates in part with the Upper Roan Subgroup in Zambia, and records the final stage of sedimentation during the initial rift phase.

There have been limited detrital zircon age studies of the Lower Roan Group, with the bulk of data indicating derivation from 2.0 to 1.8 Ga basement rocks (e.g., Steven and Armstrong, 2003). However, Halpin and Selley (2010) identified a small number of zircons from the lower (one grain) and upper levels (two grains) of this package, with near concordant ages of ~840 and 790 Ma, respectively, which potentially restrict the first rift phase to a ~50 m.y. period (Digital App. Fig. A2; Table A1).

The onset of the second rift phase was marked by an abrupt transition from platformal carbonate-facies rocks to largely subwave-base, locally reducing, siltstone-facies rocks of the Mwashya Subgroup (uppermost part of the Roan Group; Fig. 2). Coeval mafic magmatism is recorded by thin intervals of pyroclastic rocks within the basin depocenter (Lefèbvre, 1985), extrusive rocks dated at ~765 Ma along the western basin margin (Kampunzu et al., 2000; Key et al., 2001), and widespread yet volumetrically minor mafic intrusive rocks emplaced principally at the level of Upper Roan Subgroup evaporites. The overlying basal Nguba Group interval is a distinctive variably reducing glaciogenic diamictite, the Grand Conglomerate, which is conventionally equated with the first globally recognized Cryogenian glacial phase at ~740 Ma (Batumike et al., 2007). During the Mwashya Subgroup-Grand Conglomerate phase of deposition, there was a progressive widening of the depocenter, such that successively younger strata lap onto basement at the basin periphery (e.g., Kibaran and southern Kafue anticline inliers). For example, an oxidizing arenaceous facies variant of the Mwashya Subgroup, partly hosted in laterally discontinuous depocenters, unconformably overlies basement at the immediate periphery of the Kibaran inlier (Broughton and Rogers, 2010; Schmandt et al., 2013; Kennedy et al., 2018). The equivalent sequence overlies probable Dipeta Subgroup strata ~10 km to the south (Kennedy et al., 2018), indicating that the onset of northward-propagating accommodation development during the second rift stage may have been diachronous (Fig. 2).

Lateral thickness variation is pronounced within the Nguba and Kundelungu Groups (Fig. 2). This package, which includes a second diamictite at the base of the Kundelungu Group, chronostratigraphically correlated with ~635 Ma global glaciation, comprises a mixed succession of dominantly shallow water and subaerial carbonate and siliciclastic strata (Batumike et al., 2007). Onset of basin inversion was considered by Batumike et al. (2007) to have occurred during basal Kundelungu Group sedimentation; however, it is more widely accepted that upper Kundelungu Group levels, specifically an interval known as the Ngule Subgroup (Fig. 2; Digital App. Table A2), records the transition to basin closure. Uppermost Katangan Supergroup strata (Biano Group; Digital App. Table A2), preserved to the north of the Central African Copperbelt, contain a youngest detrital muscovite 40Ar/39Ar age population of 573 ± 5 Ma, interpreted by Master et al. (2003) to indicate foreland basin sedimentation.

Basin inversion and orogenesis

The maximum permissible age of orogenesis is constrained to ~600 to 560 Ma from ages of metamorphism and probable synorogenic intrusive activity peripheral to or near the southern margin of the Katangan basin, respectively (Hanson et al., 1993; John et al., 2004). The majority of the basin underwent greenschist- to subgreenschist-grade metamorphism. Higher metamorphic grades are recorded in the Domes region, where they were synchronous with the earliest phase of structural fabric development (Cosi et al., 1992).

Published studies present differing peak metamorphic conditions for the Domes region. Studies of rocks from the deepest exposed levels in the structural pile (e.g., basement and structurally interleaved or immediately overlying Lower Roan) suggest temperatures of ~750°C and pressures of ~13 kbar (by Cosi et al., 1992; John et al., 2004; Eglinger et al., 2016); however, the results of garnet-biotite-plagioclase-quartz and garnet-aluminosilicate-quartz-plagioclase geobarometry studies at the Sentinel deposit, on the southeastern flank of the Kabompo dome, suggest significantly lower pressures of ~9 to 10 and ~7 kbar, respectively (Meighan, 2015). Zircon-in-rutile and garnet-biotite geothermometry results from Sentinel also suggest lower peak metamorphic temperatures up to (and possibly exceeding) 550° to 600°C (Meighan, 2015).

Burial to depths of >50 km required for the highest reported pressure estimates in the Domes region are difficult to reconcile with the apparent intracratonic setting and known stratigraphy preserved in the region. Even the estimates of 7 kb are difficult to reconcile with the regional geology in which the Domes region is surrounded to both the north and south by age equivalent, weakly metamorphosed rocks. Some workers have questioned the robustness of pressure estimates from metasomatized rocks (e.g., strongly Mg-metasomatized quartz-talc-kyanite white schists), which likely contained highly saline and/or CO2-rich pore fluids (e.g., Meighan et al., 2012; Zimba, 2012). Nonetheless, the available data indicate most of the Domes region was heated to over 500°C, with at least some parts reaching temperatures above 600°C. Synorogenic intrusions have been recognized locally in the Domes region (First Quantum Minerals, unpub. data, 2010); however, in the absence of more compelling evidence for widespread magmatic heat input at this time, significant tectonic burial seems the most likely explanation for the attainment of peak metamorphic conditions. These controversies notwithstanding, the peak metamorphism is reasonably well constrained to the period 545 to 530 Ma by the U-Pb dating of peak metamorphic monazite (Steven and Armstrong, 2003; John et al., 2004; Turlin et al., 2016). A history of progressive exhumation and cooling, likely in part extensionally driven, is recorded in association with amphibolite and ultimately greenschist facies assemblages between 525 and 470 Ma (Cosi et al., 1992; John et al., 2004; Eglinger et al., 2016; Turlin et al., 2016).

In the northern areas of the Kafue anticline, peak metamorphic grades are mainly recorded by greenschist facies assemblages (biotite and/or phlogopite), which typically also define a single penetrative foliation. Slightly higher metamorphic grades are recorded by amphibole-bearing assemblages in outliers of Katangan Supergroup strata within the southern part of the Kafue anticline (Mendelsohn, 1961). A progressive decrease in metamorphic grade northward into the Democratic Republic of Congo was observed by François and Cailteux (1981). Here, mid- to lower-greenschist facies assemblages dominate (chlorite, muscovite, and/or phengite), and while prominent growth alignment of micas is not always developed, coarse-grained sedimentary facies commonly possess flattened and elongated clasts (e.g., Schmandt et al., 2013). There are no direct age constraints on the timing of peak metamorphism in either the Kafue anticline or Democratic Republic of Congo.

The shape of the orogen is defined by a convex-northward array of folds and reverse faults, most clearly shown by the curvilinear outcrop patterns of Roan Group strata in the Democratic Republic of Congo (Fig. 1). This geometry, akin to oroclinal bending, coupled with the decrease in metamorphic grade north of the Domes region, has conventionally been interpreted to indicate net northward tectonic transport (e.g., Kampunzu and Cailteux, 1999). Thrust duplication of the stratigraphy across extensive flats is most compelling close to the northern margin of the basin, underpinning models that invoke allochthoneity and tectonic telescoping of major portions of the Katangan Supergroup (A.P. François, unpub. report, 1973; François and Cailteux, 1981; Porada and Berhorst, 2000; Jackson et al., 2003).

Deposit types and regional metal zoning

Most of the metal in the Central African Copperbelt is located in a ~250-km-long arcuate belt projecting between the Kafue anticline in the southeast and the Kibaran block in the northwest (Fig. 1). These ores are typically stratiform, in many instances possessing kilometer-scale strike lengths. Most are hosted by reducing elements of the Lower Roan Subgroup (classic Cu-Co ores), but there are subtle differences in stratigraphic position and facies type (Fig. 2) and significant contrast in structural style between ores in Zambia and those in the Democratic Republic of Congo; for these reasons, the belt is conventionally subdivided into the Zambian and Congolese Copperbelts. Classic Zambian Copperbelt ores occur either within reducing facies of the Copperbelt Orebody Member (argillite-hosted ores) or enclosing arenaceous strata (arenite-hosted ores), the latter having locally contained preorestage hydrocarbon accumulations (Annels, 1979; Selley et al., 2005). The deposits are moderately to strongly folded, locally decoupled from their footwalls along low magnitude thrusts but otherwise preserve their original stratabound character (Mendelsohn, 1961; Molak, 1995; McGowan et al., 2003). Classic Congolese Copperbelt ores are hosted by carbonates and argillites of the lower Mines Subgroup, possessing considerably more complex, high-strain, and locally chaotic geometries that in part reflect halokinetic processes (Hitzman et al., 2012). Several vein- and/or fracture-controlled Cu-Co deposits involve a broad spatial association with classic Mines Subgroup-hosted ores but typically occur at Mwashya Subgroup levels (Taylor et al., 2013). Although the Congolese Copperbelt is effectively contiguous with the Zambian Copperbelt, the largest of the classic deposits, Kolwezi and Tenke, are located in the northwestern part of the belt.

At the extremities of the Zambian and Congolese Copperbelts, stratiform Cu ores occur at the base of the Grand Conglomerate (Kamoa and Fishtie; Figs. 1, 2). These deposits are located in sequences with abnormally condensed or absent Roan Group footwalls, and the ores are relatively weakly deformed (Broughton and Rogers, 2010: Hendrickson et al., 2015). The form of the supergiant Kamoa group of deposits in particular contrasts greatly with complex strain patterns of the neighboring Kolwezi system.

Within the inner arc of the Congolese Copperbelt, there is a progressive southward change in metal association from Cu-Co to U-Ni-Co (Au-platinum group element [PGE]) and Pb-Zn-(Cu-Ge) (Fig. 1). The belt of U-dominant ores, of which Shinkolobwe is the largest, are hosted in lower Mines Subgroup strata and possess the same complex structural configurations as the bounding Cu-Co ores (Derriks and Vaes, 1956). The largest of the Pb-Zn deposits, Kipushi, is spatially associated with fragmental Mines Subgroup strata but occurs in lower Nguba Group carbonate-rich rocks and is hosted by a system of fault zones and veins (Intomale and Oosterbosch, 1974; Heijlen et al., 2008) that are most recently interpreted as having formed along a depositional escarpment (Turner et al., 2018).

Deposits in the Domes region are varied in their metal content, structural style, and stratigraphic position. The Lumwana Cu-U system (Fig. 1) is hosted in a series of multiply reactivated basement-hosted shear zones, which project to the basal Katangan Supergroup contact in the case of the Malundwe deposit (Bernau et al., 2013). The stratiform Enterprise Ni deposit (Capistrant et al., 2015) is hosted in quartz-, carbonate-, and carbon-rich metasedimentary Lower Roan Subgroup rocks containing abundant kyanite, talc, and magnesian chlorite. The Sentinel Cu-(Ni) deposit is hosted by kyanite-bearing carbonaceous phyllite, which has been variously correlated with the Copperbelt Orebody Member (Steven and Armstrong, 2003) and the Mwashya Subgroup (Hitzman et al., 2012). In each of these systems, the earliest phase of sulfide growth is contained within the S1 foliation and/or included in peak metamorphic porphyroblasts (Steven and Armstrong, 2003; Capistrant et al., 2015; Turlin et al., 2016). The Kansanshi Cu-(Au-Mo) occurs within a tightly folded sequence of lower Nguba Group strata (including Grand Conglomerate) and differs from other Dome region deposits in that it is controlled by late-stage brittle structures, centimeter- to meter-scale veins, and/or their broad alteration selvages (Torrealday et al., 2000; Broughton et al., 2002; Hitzman et al., 2012). A similar vein-controlled Cu deposit, Frontier, occurs on the eastern flank of the Kafue anticline, where it is hosted by shales of the Mwashya Subgroup (Hitzman et al., 2012).

Deposit geochronology

A near-complete summary of isotopic ages from Central African Copperbelt ores is given in Sillitoe et al. (2017a). The reader is referred to this publication for detail, and a summary diagram is reproduced herein (Fig. 3).

Fig. 3.

Summary of published isotopic ages for Cu ± Co, U ± Ni-Co, and Pb-Zn-Cu deposits and prospects of the Central African Copperbelt. Largely reproduced from Sillitoe et al. (2017a), including references therein. Kipushi age data from Schneider et al. (2007). Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt,

Fig. 3.

Summary of published isotopic ages for Cu ± Co, U ± Ni-Co, and Pb-Zn-Cu deposits and prospects of the Central African Copperbelt. Largely reproduced from Sillitoe et al. (2017a), including references therein. Kipushi age data from Schneider et al. (2007). Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt,

A variety of isotopic geochronometers involving ore- and gangue-sulfide/oxide have been used in attempts to constrain timing of metal introduction: U-Pb uraninite/brannerite

(Darnely et al., 1961; Cahen et al., 1971; Meneghel, 1981; Loris et al., 1997; Decree et al., 2011), Re-Os Cu-Co sulfide (F. Barra and D. Broughton, reported in Selley et al., 2005; Muchez et al., 2015), Re-Os molybdenite (Torrealday et al., 2000; Barra et al., 2004; Nowecki, 2014; Sillitoe et al., 2015, 2017a), Re-Os Zn-Cu-Ge sulfide (Schneider et al., 2007), and Pb-Pb Cu sulfide (Richards et al., 1988). Of these, Re-Os molybdenite is demonstrably most robust, with consistently low error ranges and tightly bracketed synorogenic ages (largely postpeak metamorphic) of 547 to 480 Ma. A similarly well-constrained Re-Os Zn-Cu-Ge age of ~450 Ma has been determined for at least one stage of mineralization at Kipushi (Schneider et al., 2007).

The U-Pb uraninite/brannerite geochronometer produces slightly larger errors but in most cases accords with Re-Os molybdenite data from similar deposit types. Again, postpeak metamorphic ages dominate the data set, but it also generates Cryogenian ages for some of the Congolese Copperbelt deposits. Of these, the ~650 Ma age for Shinkolobwe is most robust (Decree et al., 2011) and perhaps most significant in that, with the exception of a small arenite-hosted Zambian Copperbelt deposit (Mindola), is the only system where the U resource was economic. A single Pb-Pb analysis of disseminated Cu sulfides from the Musoshi classic Cu-Co deposit, northern Zambian Copperbelt, also returns a Cryogenian age (Richards et al., 1988), occurring within error of that for Shinkolobwe.

Rhenium-Os Cu-Co sulfide geochronology is by far most ambiguous, producing large errors and unusually old ages (~820–815 Ma), unrepeated by other geochronometers. The difficulty in interpreting the Re-Os Cu-Co sulfide ages is frustrating, in that apart from uraninite from Shinkolobwe and sphalerite-bornite from Kipushi, it is the only geochronometer directly related to a principal ore phase. While these relatively old ages should be treated with caution, it must also be noted that they are geologically reasonable. The ~820 to 815 Ma age is recorded only in classic stratiform Lower Roan Subgroup-hosted ores, conforming to the permissible age range of host strata and thus suggestive of diagenetic fluid input (e.g., Muchez et al., 2015). Similarly, a pre- to early orogenic isochron age of ~583 Ma for sulfide growth in the Zambian Copperbelt is consistent with S isotope evidence of sour gas replacement in some of the deposits, a condition that may indicate significant burial (i.e., >4 km) prior to Cu introduction (Selley et al., 2005).

By contrast, the more robust Re-Os molybdenite geochronometer generates ages that are, in several cases, difficult to reconcile with other geologic criteria. For every deposit type, diverse in their stratigraphic and structural characteristics, to have formed near simultaneously is concerning but not unreasonable. However, the near-ubiquitous, postpeak metamorphic age range is very difficult to reconcile with the fact that stratiform Zambian Copperbelt ores are macroscopically deformed and metamorphically recrystallized, whereas the grain-scale relationships of sulfide and metamorphic fabrics at Sentinel, Enterprise, and Lumwana indicate that at least some phase of Cu introduction was pre-synpeak metamorphic. For the Domes region deposits at least, unless the robustness of the U-Pb monazite geochronometer (source of the peak metamorphic constraint) is put into question, the Re-Os molybdenite ages must be recording only part of the mineralizing history. This argument is less certain for the Zambian Copperbelt ores, as there are no precise ages of metamorphism in this region, and synchronicity with the Domes region is by inference only; i.e., exhumation of the Domes region could have been coeval with peak metamorphism in the Zambian Copperbelt.

A postpeak metamorphic age for Kansanshi (~512 and 503 Ma; Darnley et al., 1961; Torrealday et al., 2000) is, however, entirely consistent with fabric relationships. The host vein array clearly postdates the principal phase of folding and crosscuts the youngest fabric in the rock.

In summary, the data appear to record multiple phases of metal-bearing fluid flow, from the onset of basin growth until at least 70 m.y. after peak metamorphism in the Domes region: ~815, 650, 585, 535, 515, 500, and 450 Ma. It is not possible to constrain the relative volume of metal introduced during each phase beyond Mo having been syn- to principally postpeak metamorphic and Zn-Cu-Ge introduction having occurred during the Ordovician.

Structural Style

The Katangan basin possesses significant vertical and lateral variations in structural style. A complex arrangement of high-amplitude, typically noncylindrical, upright to recumbent macroscopic folds, layer-subparallel to highly layer-oblique decoupling surfaces, records progressive deformation throughout basin growth and subsequent closure. Near-basin-wide decoupling zones occur at a number of levels in the lower and middle portions of the Katangan Supergroup. In all but the extreme peripheries of the basin, these are principally located within Roan Group strata and are typically associated with breccia facies (Kampunzu and Cailteux, 1999; Porada and Berhorst, 2000; Jackson et al., 2003; Selley et al., 2005; Hitzman et al., 2012; Woodhead, 2013).

Many of the strain patterns are attributable to halokinesis and vary in accordance with the original volume of salt that accumulated in different parts of the basin (Jackson et al., 2003). The consideration of data below involves the entirety of the Katangan basin but focuses on relationships in the Congolese Copperbelt.

Zambian Copperbelt

In the Zambian Copperbelt, a main zone of decoupling occurs in the presently carbonate-dominated succession of the Upper Roan Subgroup (Fig. 4). Additional layer-subparallel fault zones are developed at stratigraphic interfaces with abrupt rheologic contrast, including the margins of the Copperbelt Orebody Member, and affect the configuration of some deposits (e.g., Nchanga; McGowan et al., 2003). Except for localized recumbent folds that root into the decoupling zone on the eastern flank of the Kafue anticline (e.g., Frontier; Hitzman et al., 2012), there is no demonstrable thinskinned duplication of preserved stratigraphy above the level of the breccias. Mwashya Subgroup, as well as Nguba and rarely preserved intervals of lower Kundelungu Groups strata, exhibit a nonfragmental, macroscopically layer-cake form (i.e., nonhalokinetic) and have been deformed in conjunction with breccia facies and underlying Lower Roan Subgroup strata by high-wavelength/amplitude, generally upright, basement-involved folds. Folds are doubly plunging, west-northwest–east-southeast to north-northwest–south-southeast, and associated with a single penetrative, predominantly phlogopite-defined foliation. These geometric relationships imply there was limited postpeak metamorphic lateral displacement on the decoupling zones and, perhaps more equivocally, that they nucleated prior to macroscopic folding.

Fig. 4.

Geologic map of the Zambian Copperbelt (modified from Selley et al., 2005). With the exception of the Mwashya Subgroup-hosted Frontier deposit, all major deposits are hosted within Lower Roan Subgroup strata. A regional decoupling surface, defined by breccia facies and deformed by upright basement-involved folds, is principally positioned in Upper Roan Subgroup strata. On the western flank of the Kafue anticline, the decoupling surface cuts down-section to middle levels of the Lower Roan Subgroup. Interpolation of the points at which the decoupling surface changes from flat to ramp configuration defines a northwest trace that projects between Chibuluma West and Musoshi and corresponds to the western limits of known Copperbelt Orebody Member strata. Recumbent folds occur above the decoupling zone in the region of the Frontier deposit. Abbreviations: CCB = Congolese Copperbelt, Gp(s) = Group(s), Sgp(s) = Subgroup(s), ZCB = Zambian Copperbelt.

Fig. 4.

Geologic map of the Zambian Copperbelt (modified from Selley et al., 2005). With the exception of the Mwashya Subgroup-hosted Frontier deposit, all major deposits are hosted within Lower Roan Subgroup strata. A regional decoupling surface, defined by breccia facies and deformed by upright basement-involved folds, is principally positioned in Upper Roan Subgroup strata. On the western flank of the Kafue anticline, the decoupling surface cuts down-section to middle levels of the Lower Roan Subgroup. Interpolation of the points at which the decoupling surface changes from flat to ramp configuration defines a northwest trace that projects between Chibuluma West and Musoshi and corresponds to the western limits of known Copperbelt Orebody Member strata. Recumbent folds occur above the decoupling zone in the region of the Frontier deposit. Abbreviations: CCB = Congolese Copperbelt, Gp(s) = Group(s), Sgp(s) = Subgroup(s), ZCB = Zambian Copperbelt.

While evidence of thrust duplication is lacking, a major breccia-associated ramp is clearly developed on the western margin of the Kafue anticline. The ramp progressively cuts down-section, in a northwestward direction, from Upper Roan Subgroup levels to the base of the Copperbelt Orebody Member (Fig. 4). This geometry leads to juxtaposition of Upper Roan Subgroup carbonate-dominant strata and relatively coarse grained Lower Roan Subgroup synrift strata in the hanging walls of the Chibuluma and Chibuluma West orebodies, and lateral truncation of ore hosted in the Copperbelt Orebody Member at Musoshi. There are no documented kinematic constraints for ramp displacement, but two scenarios could account for the observed geometric relationships. First, the ramp nucleated during a phase of extensional collapse and preserved net normal displacement throughout inversion. The cartoon in Figure 5a shows the ramp linking basinward into an Upper Roan Subgroup flat, downthrown to a deep structural level by a basement-involved growth fault. Alternatively, compressional reactivation of a similar fault system could have produced the observed geometric relationships, the apparent removal of stratigraphy across the ramp resulting from shortcutting of an uplifted fault block from a deep-seated flat (Fig. 5b). Both scenarios are problematic in that some duplication of stratigraphy is implied but not observed, i.e., basinward in the extensional model and toward the Kafue anticline in the case of compression. Possible explanations include paucity of drilling downdip of the ramp and erosion over the crest of the Kafue anticline, respectively. The development of recumbent décollement folds on the eastern flank of the Kafue anticline, upon which upright folds appear superimposed, and the development of prefoliation stratiform vein extensional arrays in some deposits (e.g., Annels, 1989; Brems et al., 2009) may favor the compressional model. A compressional history does not, however, preclude a prior phase of extension.

Fig. 5.

Schematic cross sections illustrating alternative hypotheses for the displacement history of the Zambian Copperbelt decoupling surface. (a) Extensional collapse, with the down-section-cutting ramp on the western flank of the Kafue anticline connecting to a flat positioned at deeper structural levels. (b) Northeast-directed thrusting during the earliest stages of basin inversion, the western ramp representing a short-cut from deeper structural levels. The asymmetry of recumbent folds is consistent with broadly NE-directed thrusting. In both scenarios, localized duplication of the stratigraphy is required. Abbreviations: Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 5.

Schematic cross sections illustrating alternative hypotheses for the displacement history of the Zambian Copperbelt decoupling surface. (a) Extensional collapse, with the down-section-cutting ramp on the western flank of the Kafue anticline connecting to a flat positioned at deeper structural levels. (b) Northeast-directed thrusting during the earliest stages of basin inversion, the western ramp representing a short-cut from deeper structural levels. The asymmetry of recumbent folds is consistent with broadly NE-directed thrusting. In both scenarios, localized duplication of the stratigraphy is required. Abbreviations: Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Domes region

In addition to the high grade of metamorphism in the Domes region, the intensity of deformation is considerably greater than in other parts of the Katangan basin. A complex history, including basement-involved thrusting, polyphase folding, and multiple fabric development, is best documented at the deposit scale (Cosi et al., 1992; Broughton et al., 2002; Steven and Armstrong, 2003; Bernau et al., 2013; Capistrant et al., 2015; Turlin et al., 2016) but also revealed in regional studies (Key et al., 2001; Barron, 2003). These studies reveal an early formed, thin-skinned duplex-type arrangement, consisting of a lower structural level of basement-involved thrusts and recumbent to inclined folds and an upper level of locally high amplitude recumbent folds (some of which are bounded by breccias) and thrusts that root into an intervening décollement (Fig. 6). In the region of the Solwezi dome (Fig. 1), Barron (2003) places the decoupling surface in the Upper Roan Subgroup, with recent exploration having demonstrated that this position is occupied by evaporite-related breccias similar to those of the Zambian Copperbelt. Decoupling immediately above the Lower Roan Subgroup is also recorded at the Enterprise Ni deposit, located on the eastern flank of the Kabompo dome, where a stratabound body of breccia facies apparently stopes out Upper Roan Subgroup strata (Capistrant et al., 2015).

Fig. 6.

Schematic representation of the fold and fault configuration of the Domes region. The Lumwana deposits, Malundwe and Chimiwungo, are positioned along low-angle thrusts, locally refolded by shallowly inclined F2 closures. Enterprise and Sentinel are interpreted to be positioned below a breccia-defined decoupling surface at Upper Roan Subgroup levels. Kansanshi occurs at a supradetachment level, overlying a northward-thickening, synrift package. Major shortening phases were punctuated by episodes of extensionally driven exhumation. Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 6.

Schematic representation of the fold and fault configuration of the Domes region. The Lumwana deposits, Malundwe and Chimiwungo, are positioned along low-angle thrusts, locally refolded by shallowly inclined F2 closures. Enterprise and Sentinel are interpreted to be positioned below a breccia-defined decoupling surface at Upper Roan Subgroup levels. Kansanshi occurs at a supradetachment level, overlying a northward-thickening, synrift package. Major shortening phases were punctuated by episodes of extensionally driven exhumation. Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Basement-involved deformation below the décollement is heterogeneously developed about at least some dome margins and most extreme at the Lumwana Cu-U deposits, located in the core of the Mwombezhi dome (Figs. 1, 6). Here, highly strained lower Katangan Supergroup strata (i.e., Lower and Upper Roan Subgroups) occur structurally interleaved with basement granitic gneisses, across a combination of shear zones and isoclinal fold closures (Cosi et al., 1992; Bernau et al., 2013). The structural model of Cosi et al. (1992) invokes an alpine-type geometry for the domes, involving thin-skinned, rootless basement-cored nappes. While this geometry is permissible at the deposit scale, regional gravity data indicate that the domes correspond with domains of anomalously low mass (Digital App. Fig. A1; Duffett et al., 2010), suggesting that significant volumes of low-density granitic basement material persist to depth. Thus, while isoclinal folds potentially possess amplitudes in the order of 5 km, they are not rootless and probably nucleated at perturbations along the basement-Lower Roan Subgroup interface.

Lumwana mineralization is hosted in a ~50-m-wide subhorizontal shear zone, which possesses a composite schistose fabric. Early peak- and postpeak metamorphic kyanite + biotite fabrics (S1 and S2, respectively) are overprinted by retrograde muscovite and associated S3 shear fabrics. The latter possess top-to-the-north kinematic indicators and were interpreted by Turlin et al. (2016) to record dome exhumation. In neighboring Lower Roan Subgroup strata, two generations of kyanite are also observed, the second associated with a recumbent isoclinal folding event (F2). Evidence exists for partial replacement of S1 kyanite prior to growth of S2 kyanite, suggesting an intervening episode of instability, possibly related to decompression. Collectively, the fabrics and their associated metamorphic assemblages appear to record two phases of shortening-related burial, separated and postdated by pulses of exhumation. We consider the exhumation phases best explained by crustal extension, an interpretation most convincingly recorded by the low-angle retrograde S3 shear fabric, and that extension appears the most efficient means of returning once deeply buried rocks to high crustal levels.

Two penetrative foliations are also locally recorded in Katangan Supergroup strata from above the regional décollement, an early layer-parallel fabric associated with highamplitude, shallowly inclined to recumbent folds, and an S2 crenulation associated with upright to inclined folds (Barron, 2003). In contrast to the Mwombezhi dome fabrics, peak metamorphic conditions appear to be recorded by S2 biotite + garnet + hornblende + kyanite (Barron, 2003). Layer-parallel S1 is defined as an apparent greenschist facies assemblage of muscovite + biotite, a paradox that may be explained a post-S1, pre-S2 phase of retrogression, similar to that recorded at Lumwana. At the Kansanshi deposit, a single penetrative layer-parallel foliation (S1) is overprinted by at least two generations of inclined to upright folds that locally involve the development of crenulation cleavages but apparently lack associated metamorphic mineral growth; mineralized veins are interpreted to have formed synchronously with the latest folding phase (Broughton et al., 2002).

Southern and central Congolese Copperbelt

A prominent decoupling surface projects to surface in the southeastern limit of the Congolese Copperbelt, about the fringe of the Luina dome (Fig. 7). The geometry and stratigraphic position of the decoupling surface has similarities to that of the adjacent Zambian Copperbelt. It is broadly stratabound and separates a footwall succession of Lower Roan Subgroup syn- to postrift clastics (350–900 m) that includes a locally deposited but typically poorly Cu mineralized Copperbelt Orebody Member equivalent (5–15 m), from a hanging-wall succession of schist, fine sandstone, and carbonate, that equates to the Upper Roan Subgroup (A.P. François, unpub. report, 1958; Lefèbvre and Tshiauka, 1986). The latter locally contains reducing intervals toward its base, that have been alternatively correlated with the Mines Subgroup (Lefèbvre and Tshiauka, 1986; Woodhead, 2013) or parts of the overlying Dipeta Subgroup (A.P. François, unpub. report, 1958). In contrast to the Zambian Copperbelt, however, fold styles above and below the decoupling surface are very different (Fig. 7); subdetachment strata are deformed sympathetically with basement to produce high-wavelength, doubly plunging folds, whereas the hanging-wall succession is affected by high-amplitude, low-wavelength, upright décollement folds that root into the detachment (A.P. François, unpub. report, 1958).

Fig. 7.

Map of the Luina dome region: transition between the Zambian and Congolese Copperbelts (based on A.P. François, unpub. report, 1958; Gecamines 1:20,000 map set). To the south of the dome, the breccia-defined decoupling surface is deformed by high-wavelength, basement-cored anticlines. However, on the northern side of the dome, the decoupling surface forms the sole to relatively low-wavelength, high-amplitude, breccia-cored folds affecting Mines Subgroup and younger strata. The Mines Subgroup-hosted Kimpe deposit is the southernmost of the classic Congolese Copperbelt deposits. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 7.

Map of the Luina dome region: transition between the Zambian and Congolese Copperbelts (based on A.P. François, unpub. report, 1958; Gecamines 1:20,000 map set). To the south of the dome, the breccia-defined decoupling surface is deformed by high-wavelength, basement-cored anticlines. However, on the northern side of the dome, the decoupling surface forms the sole to relatively low-wavelength, high-amplitude, breccia-cored folds affecting Mines Subgroup and younger strata. The Mines Subgroup-hosted Kimpe deposit is the southernmost of the classic Congolese Copperbelt deposits. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Approximately 10 km north of the Luina dome, unequivocal Cu-mineralized Mines Subgroup units appear as megablocks within antiformal cores (i.e., Kimpe; Cailteux and Lefèbvre, 1975; Fig. 7). Accepting the stratigraphic frameworks of Cailteux et al. (1994) and Bull et al. (2011), wherein the Mines Subgroup correlates with Lower Roan Subgroup strata in the Zambian Copperbelt, the decoupling surface again appears to cut down-section away from the basement inliers, in this case toward the north, most likely to the top of the Lower Roan Subgroup synrift package.

The decoupling surface is inferred, rather than demonstrated, to project deep into the structural pile within the central parts of the Congolese Copperbelt. This inference is based partly on the northward persistence of high-amplitude upright folding affecting all stratigraphic levels from the Mines Subgroup to the Kundelungu Group but also by the general paucity of Lower Roan Subgroup synrift strata exposed within antiformal cores.

The northward transition to the Congolese Copperbelt is also accompanied by increasing volumes of breccia bodies. Containing predominantly Roan Group and lesser Nguba and Kundelungu Groups fragments, these bodies form irregular, discontinuous map patterns, which in conjunction with spatially associated fault and fold axial traces, define an array of predominantly WNW-, ENE-, and NW-trending lineaments (Fig. 8a, b). A major NW-striking fault zone that projects to ~5 km north of the Kansanshi deposit coincides with a southward reduction in the volume of breccia bodies and defines the transition between the Congolese Copperbelt and the Domes region (Fig. 9). All of the known ore deposits in the southern and central parts of the Congolese Copperbelt, regardless of their metal association, are hosted within or at the immediate peripheries of these bodies, many of them distributed along the trace of a laterally extensive NW-trending corridor, herein termed the Kakanda-Luisha fault zone (Fig. 8a). In the case of the classic Mines Subgroup-hosted Cu deposits, the macroscopic form of the ores is chaotic. Mineralized strata are deformed by variably oriented folds and faults and ultimately fragmented to occur as floating megablocks (colloquially referred to as “écailles”) with breccias that appear to range from matrix-rich (Cailteux, 1986; Fig. 10) to matrix-poor (Byrne, 2017). Despite the complexity of deformation, the internal geometry of écailles preserves the original stratabound character of ore, and textures such as truncated Cu-bearing veins clearly indicate a prefragmental timing of mineralization.

Fig. 8.

Structural framework of the central Congolese Copperbelt (modified from the Gecamines 1:20,000 map set). See inset for map locations. (a) Distribution of Roan Group strata and their geometric relationship to major faults and folds. A semitransparent mask is overlain on the northern Congolese Copperbelt to highlight features of the central Congolese Copperbelt. In the latter, Roan Group strata are ubiquitously fragmented and define narrow, irregular map patterns that follow the traces of irregular and discontinuous faults. (b) Linear corridors defined by alignment and/or intersections of breccia facies, faults, and fold axial traces. Note that many lineaments crosscut each other with minimal apparent mutual offset. Several of the lineaments project into the northern Congolese Copperbelt domain, where they coincide with major orebodies (e.g., Kakula, Tenke). (c) Gridded bedding dip angles overlain by fold axial traces, breccia facies, and lower levels of the Nguba Group. Systematic geometric and spatial relationships between the grid pattern and fold axial traces reflect inheritance arrays of salt walls and intervening salt withdrawal minibasins. (d) Seismic profile through a diapiric domain, Santos basin, offshore Brazil (from Trudgill, 2005). The arrangement of steeply dipping strata on the flanks of diapirs and low-angle dips within the adjacent minibasins are analogous to the map patterns shown in (c). Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, Congl. = Conglomerate, FZ = fault zone, Gp(s) = Group(s), NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 8.

Structural framework of the central Congolese Copperbelt (modified from the Gecamines 1:20,000 map set). See inset for map locations. (a) Distribution of Roan Group strata and their geometric relationship to major faults and folds. A semitransparent mask is overlain on the northern Congolese Copperbelt to highlight features of the central Congolese Copperbelt. In the latter, Roan Group strata are ubiquitously fragmented and define narrow, irregular map patterns that follow the traces of irregular and discontinuous faults. (b) Linear corridors defined by alignment and/or intersections of breccia facies, faults, and fold axial traces. Note that many lineaments crosscut each other with minimal apparent mutual offset. Several of the lineaments project into the northern Congolese Copperbelt domain, where they coincide with major orebodies (e.g., Kakula, Tenke). (c) Gridded bedding dip angles overlain by fold axial traces, breccia facies, and lower levels of the Nguba Group. Systematic geometric and spatial relationships between the grid pattern and fold axial traces reflect inheritance arrays of salt walls and intervening salt withdrawal minibasins. (d) Seismic profile through a diapiric domain, Santos basin, offshore Brazil (from Trudgill, 2005). The arrangement of steeply dipping strata on the flanks of diapirs and low-angle dips within the adjacent minibasins are analogous to the map patterns shown in (c). Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, Congl. = Conglomerate, FZ = fault zone, Gp(s) = Group(s), NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 9.

Aeromagnetic data map (total magnetic intensity, reduced to the pole, first vertical derivative) of the Domes region and its transition into the southern Congolese Copperbelt. The boundary, shown as curvilinear, segmented, predominantly NNE-dipping fault zone, is defined in part by the abrupt northward increase in the areal extent of Roan Group breccia complexes and more complex noncylindrical fold patterns. In this region, breccia facies has a distinctive high-frequency/amplitude magnetic texture, due to significant volumes of dismembered mafic intrusives. Map location shown in inset. Abbreviations: CCCB = central Congolese Copperbelt, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 9.

Aeromagnetic data map (total magnetic intensity, reduced to the pole, first vertical derivative) of the Domes region and its transition into the southern Congolese Copperbelt. The boundary, shown as curvilinear, segmented, predominantly NNE-dipping fault zone, is defined in part by the abrupt northward increase in the areal extent of Roan Group breccia complexes and more complex noncylindrical fold patterns. In this region, breccia facies has a distinctive high-frequency/amplitude magnetic texture, due to significant volumes of dismembered mafic intrusives. Map location shown in inset. Abbreviations: CCCB = central Congolese Copperbelt, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 10.

Cross section through the Kambove deposit (refer to Fig. 8 for location), showing the typical chaotic structural geometry of the Mines Subgroup-hosted ores of the central Congolese Copperbelt (after Cailteux, 1986). The breccia complex and its carapace of Mwashya Subgroup strata have been emplaced to the level of the Ngule Subgroup. The level of ore is highlighted in the thin red traces, demonstrating preservation of stratabound form, despite the structural complexity. Abbreviations: Gp = Group, undiff. = undifferentiated.

Fig. 10.

Cross section through the Kambove deposit (refer to Fig. 8 for location), showing the typical chaotic structural geometry of the Mines Subgroup-hosted ores of the central Congolese Copperbelt (after Cailteux, 1986). The breccia complex and its carapace of Mwashya Subgroup strata have been emplaced to the level of the Ngule Subgroup. The level of ore is highlighted in the thin red traces, demonstrating preservation of stratabound form, despite the structural complexity. Abbreviations: Gp = Group, undiff. = undifferentiated.

Breccia bodies are in most cases positioned within the cores of anticlines but also occur along steeply dipping faults that cut stratigraphy at high angles and along the hanging walls of thrust flats emplaced at a specific stratigraphic interval of the Kundelungu Group (Ngule Subgroup; Figs. 2, 10). The latter appear in map view as small isolated klippen, their footwall thrust surfaces rooting into neighboring antiformal cores and exhibiting limited strike extent, implying low-magnitude reverse displacements. In the case of Kambove (Fig. 10), the ore-bearing breccia complex is apparently emplaced into the Ngule Subgroup level, along with a preserved carapace of Mwashya and Grand Conglomerate strata, its footwall contact possessing a distinctly curvilinear cross-sectional trace. Rare examples of breccia, typically dominated by Kundelungu Group material, occupy the cores of tight high-wavelength synclines, their bases apparently parallel to underlying Ngule Subgroup strata (François, 1986). While similar to the mineralized klippen in terms of their stratigraphic position, perhaps marking the position of a once laterally extensive breccia complex near the top of the Kundelungu Group, the synformhosted breccia bodies are barren and preserve no evidence of Nguba Group carapaces.

The arrangement of breccia-defined lineaments demonstrates that the arcuate form of the Congolese Copperbelt results not from oroclinal bending, but the complex interplay of variously oriented segments. Moreover, it can be seen from Figure 8b that several of the more extensive lineaments (up to 100 km) crosscut each other without apparent mutual offset, indicating that they have not accommodated significant (i.e., tens of kilometers) displacement.

Evidence of diapirism: A halokinetic origin for the breccia bodies was first proposed by de Magnee and François (1988) based on observations at the Kipushi deposit and subsequently extended to the entire Congolese Copperbelt in the sophisticated structural model of Jackson et al. (2003). The latter, in particular, drew analogies with map patterns observed in modern and ancient halokinetic basins and invoked a complex history of salt-wall emplacement during basin growth, ultimately welded and breached to source an extensive, extrusive salt allochthon, overridden by far-traveled (>65 km) thrust sheets during orogenesis.

Clear evidence of a link between brecciation and diapir amplification during basin growth is shown by geometric and stratal thickness changes on the margins of the breccia bodies. The example given in Figure 11 shows dramatic thinning of lower and middle Nguba Group strata onto the immediate flanks of the Kipese breccia complex. Onset of diapir amplification is recorded by the Grand Conglomerate, which thins laterally from a maximum of >1,500 m in the Tantara valley to just a few tens of meters on the western and eastern flanks of the breccia complex. A syndepositional down-building mechanism (e.g., Fuchs et al., 2011) was likely, where the diapir head initiated with, and maintained, a near-surface position, while sediment accumulated in peripheral minibasins whose subsidence was largely driven by lateral migration of underlying salt toward the diapir stem. Rift-related tectonism recorded by upper Mwashya Subgroup strata, and/or abrupt differential loading by rapidly deposited glacial outwash during Grand Conglomerate deposition, provided the likely triggers for destabilization of Roan Group salt.

Fig. 11.

Map of the Kipese-Shinkolobwe region, central Congolese Copperbelt, highlighting the complex configuration of brecciated Roan Group strata and thickness variation at Nguba Group levels, the latter most evident in the Tantara valley. The map patterns represent a rarely exposed profile through a minibasin, with thinning of Grand Conglomerate strata in particular onto the flanks of the former Kipese diapir. Note also the transgressive nature of the breccia facies, cutting upsection to Kundelungu Group stratigraphic levels. Refer to Digital Appendix Table A2 for a detailed stratigraphic framework for the Nguba and Kundelungu Groups. Modified from the Gecamines 1:20,000 map set. Abbreviations: CCCB = central Congolese Copperbelt, Congl. = Conglomerate, Gp = Group, NCCB = northern Congolese Copperbelt, Sgp = Subgroup, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 11.

Map of the Kipese-Shinkolobwe region, central Congolese Copperbelt, highlighting the complex configuration of brecciated Roan Group strata and thickness variation at Nguba Group levels, the latter most evident in the Tantara valley. The map patterns represent a rarely exposed profile through a minibasin, with thinning of Grand Conglomerate strata in particular onto the flanks of the former Kipese diapir. Note also the transgressive nature of the breccia facies, cutting upsection to Kundelungu Group stratigraphic levels. Refer to Digital Appendix Table A2 for a detailed stratigraphic framework for the Nguba and Kundelungu Groups. Modified from the Gecamines 1:20,000 map set. Abbreviations: CCCB = central Congolese Copperbelt, Congl. = Conglomerate, Gp = Group, NCCB = northern Congolese Copperbelt, Sgp = Subgroup, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

At a more regional scale, diapiric form is revealed in fold patterns. Figure 8c shows a map of gridded bedding dip magnitude. While complex at first glance, it can be seen that the inclination of layering varies systematically with map patterns of both breccia bodies and fold axial traces. There is a relatively consistent relationship between domains of high dip and outcropping lower Nguba Group and older stratigraphic levels, the latter principally, but not exclusively, occupying cores of high-amplitude, partly dismembered, highly noncylindrical antiforms. By contrast, the intervening synforms, cored by middle Nguba Group and younger strata, typically correspond with domains of anomalously low bedding dip. This arrangement of tight, typically breccia-cored antiforms and neighboring gentle, high-wavelength synforms is interpreted to be inherited from a 3-D network of preorogenic salt walls and intervening salt withdrawal minibasins (cf. Fig. 8d). In this sense, the bedding dip grid can be visualized as a proxy for the topography of the upper surface of salt, i.e., whites and reds in Figure 8c defining topographic highs on the crests and upper flanks of diapirs, blues and purple corresponding to topographic lows at the bases of salt withdrawal minibasins.

Regional isopach maps demonstrate a relationship between the distribution of salt walls and first-order subbasin compartments. A grid of upper Nguba Group thickness data shows this most clearly (Fig. 12a), in part due to the relatively even distribution of data points at this level. The unit thickens progressively southward of the domain of major classic ores that defined the northern arm of the copper arc (i.e., Kansuki fault zone) and southwestward of the Kakanda-Luisha fault zone. Major and abrupt thickness changes occur across the ENE-striking, breccia-lined Luankonko and Monwezi fault zones. A second-order subbasin occurs to the east of the Kakanda-Luisha fault zone (Sesa subbasin), enclosed by breccia bodies, projecting along a more diffuse west-northwest corridor, encompassing U mineralization at Shinkolobwe and Swambo. The lack of mutual offset of this trend and the Kakanda-Luisha fault zone indicates only minor displacement post Nguba Group sedimentation.

Fig. 12.

Isopachs of Nguba Group strata, overlain by macroscopic structural elements, Roan Group map patterns, and major deposits. Thickness data were derived from outcrop traverses indicated on the Gecamines 1:20,000 map set. Tabulated data and the method of thickness calculation are given in Digital Appendix Table A3 and Figure A3, respectively. Classification of data is based on the quality of parameters used to calculate unit thickness: i.e., unit outcrop, boundary constraints, and bedding orientation. Those data labeled “robust” were determined from traverses where these parameters are moderately to well constrained, with measurements of sufficient precision to examine macroscale thickness variation. Those data labeled “minimum” thickness relate to measurements made where upper or lower unit boundaries are faulted or do not project to surface. “Maximum” thickness refers to data where bedding orientation is poorly constrained, and the calculation assumes a vertical dip. Those data labeled “approximate” lack sufficient outcrop, formation boundary constraints, and/or bedding orientation to derive an accurate thickness measurement. All data are used in construction of the grids, as nonrobust classes typically vary systematically with robust data, and while they do not represent absolute thicknesses, they contribute to important macroscale geometric patterns; care should be taken where nonrobust data coincide with local anomalies in the grids. (a) Upper Nguba Group patterns reveal abrupt southward thickening across major ENE-, NW-, and WNW-striking fault zones. Several of the fault zones coincide with small elongate domains of anomalously condensed strata (red arrows). These represent areas of low-accommodation development on the crests of salt walls (cf. Fig. 8d). Classic stratiform Congolese Copperbelt ores occur mainly in condensed parts of the basin, those close to the Kakanda-Luisha fault zone along a major subbasin edge, and the larger northern deposits within a thin platform. (b) Grand Conglomerate patterns have an opposing relationship, with strata apparently thinner in the southern part of the basin system. This is an artefact of sample distribution, with most data occurring along the crests of Roan Group breccia-cored antiforms and breccia-lined fault zones. The restriction of data to salt wall crests means that the true volume of Grand Conglomerate is greatly underrepresented. The positive thickness anomaly at the fringe of the Kipese breccia to the south is more representative and corresponds to the minibasin shown in Figure 11. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 12.

Isopachs of Nguba Group strata, overlain by macroscopic structural elements, Roan Group map patterns, and major deposits. Thickness data were derived from outcrop traverses indicated on the Gecamines 1:20,000 map set. Tabulated data and the method of thickness calculation are given in Digital Appendix Table A3 and Figure A3, respectively. Classification of data is based on the quality of parameters used to calculate unit thickness: i.e., unit outcrop, boundary constraints, and bedding orientation. Those data labeled “robust” were determined from traverses where these parameters are moderately to well constrained, with measurements of sufficient precision to examine macroscale thickness variation. Those data labeled “minimum” thickness relate to measurements made where upper or lower unit boundaries are faulted or do not project to surface. “Maximum” thickness refers to data where bedding orientation is poorly constrained, and the calculation assumes a vertical dip. Those data labeled “approximate” lack sufficient outcrop, formation boundary constraints, and/or bedding orientation to derive an accurate thickness measurement. All data are used in construction of the grids, as nonrobust classes typically vary systematically with robust data, and while they do not represent absolute thicknesses, they contribute to important macroscale geometric patterns; care should be taken where nonrobust data coincide with local anomalies in the grids. (a) Upper Nguba Group patterns reveal abrupt southward thickening across major ENE-, NW-, and WNW-striking fault zones. Several of the fault zones coincide with small elongate domains of anomalously condensed strata (red arrows). These represent areas of low-accommodation development on the crests of salt walls (cf. Fig. 8d). Classic stratiform Congolese Copperbelt ores occur mainly in condensed parts of the basin, those close to the Kakanda-Luisha fault zone along a major subbasin edge, and the larger northern deposits within a thin platform. (b) Grand Conglomerate patterns have an opposing relationship, with strata apparently thinner in the southern part of the basin system. This is an artefact of sample distribution, with most data occurring along the crests of Roan Group breccia-cored antiforms and breccia-lined fault zones. The restriction of data to salt wall crests means that the true volume of Grand Conglomerate is greatly underrepresented. The positive thickness anomaly at the fringe of the Kipese breccia to the south is more representative and corresponds to the minibasin shown in Figure 11. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

A series of small areas of anomalously thin strata occur along traces of abrupt thickness gradients. Proximal to breccia bodies and indicated by red arrows in Figure 12a, these zones of anomalously condensed upper Nguba Group strata are interpreted to record limited accommodation development on diapiric crests.

Regional isopachs of the Grand Conglomerate are shown for comparison in Figure 12b, revealing curiously opposing patterns to those of the upper Nguba Group, with highly condensed domains appearing southward of the Luankonko and Monwezi fault zones. These data create an illusion of progressive southward thinning of the Grand Conglomerate, caused by the very limited distribution of data points along the fringes of breccia-cored antiforms (i.e., salt walls) and significant interpolation distances across the intervening synforms (i.e., withdrawal subbasins). The pronounced anomaly that corresponds to the thick Grand Conglomerate interval in the Tantara Valley at the margin of the Kipese breccia body is a very rare example of the base of a withdrawal minibasin exposed at the surface and is considered to provide a more accurate indication of regional thicknesses at this level.

Northern Congolese Copperbelt

The northern Congolese Copperbelt is host to the largest of the classic Mines Subgroup-hosted Cu-Co ores and the presently expanding Grand Conglomerate-hosted stratiform Cu resources of the Kamoa district (Fig. 13). It contains perhaps the most complex and contentious structural configurations in the Katangan basin. From west to east, four domains are distinguished on the basis of structural style, each with laterally transitional boundaries: Kamoa, Kolwezi, Pumpi, and Tenke domains. All domains are bounded to the south by the northward-convex trace of the Kansuki fault zone, which in terms of extensional basin configuration corresponds with an abrupt northward reduction in accommodation development, at least during upper Nguba Group sedimentation (Fig. 12a).

Fig. 13.

Structure and distribution of major Cu ± Co ores of the northern Congolese Copperbelt; semitransparent overlay masks features of the central Congolese Copperbelt. (a) Map of the northern Congolese Copperbelt, modified from the Gecamines 1:20,000 map set. The region is divided into four domains on the basis of structural style: Kamoa, Kolwezi, Pumpi, and Tenke domains. (b) Cross section B-B’ of the northern and central Congolese Copperbelt, highlighting the contrasting thin- and thick-skinned structural styles, respectively. See text for discussion. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, Congl. = Conglomerate, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, RSL = relative sea level, SCCB = southern Congolese Copperbelt, Sgp(s) = Subgroup(s), ZCB = Zambian Copperbelt.

Fig. 13.

Structure and distribution of major Cu ± Co ores of the northern Congolese Copperbelt; semitransparent overlay masks features of the central Congolese Copperbelt. (a) Map of the northern Congolese Copperbelt, modified from the Gecamines 1:20,000 map set. The region is divided into four domains on the basis of structural style: Kamoa, Kolwezi, Pumpi, and Tenke domains. (b) Cross section B-B’ of the northern and central Congolese Copperbelt, highlighting the contrasting thin- and thick-skinned structural styles, respectively. See text for discussion. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, Congl. = Conglomerate, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, RSL = relative sea level, SCCB = southern Congolese Copperbelt, Sgp(s) = Subgroup(s), ZCB = Zambian Copperbelt.

Kamoa domain: The Kamoa domain is simplest in terms of structural configuration, constituting a NE-trending belt of uppermost Roan to Kundelungu Group strata. In a similar fashion to basement dome-proximal regions in the southern Congolese and Zambian Copperbelts, the package is deformed in conjunction with unconformably underlying rocks of the Kibaran inlier into dome and basin patterns, resulting from the interference of NE-, ENE-, and WNW-trending, mainly gentle to open, upright folds (Fig. 13a). A first-order NE-trending anticline projects southwestward of the Kibaran inlier, the major deposits of Kamoa and Kakula occupying its crest and located at the intersection of second-order ENE- and WNW-trending synforms, respectively. The east-northeast trend of Kamoa projects toward Kolwezi, whereas the west-northwest trend of Kakula conforms broadly to that of the Monwezi fault zone of the central Congolese Copperbelt.

Only a single tectonic foliation is observed (Broughton and Rogers, 2010), the noncylindrical fold geometry interpreted to reflect inheritance from subbasin architecture association with Mwashya Subgroup rift stage. Lateral thickness variation is best demonstrated at the level of the Grand Conglomerate, which increases from ~180 m on the northeastern flank of the Kibaran inlier to ~1,800 m on the southern margin of the Kamoa domain (Kennedy et al., 2018). The underlying Mwashya Subgroup is represented by a proximally derived conglomerate facies, unconformably laid upon the eastern and southern flanks of the Kibaran inlier, thickening southward to include intervals of clean and argillaceous sandstone and minor carbonate (A.P. François, unpub. report, 1973; Broughton and Rogers, 2010; Schmandt et al., 2013; Kennedy et al., 2018), consistent with progressive southward increase in accommodation development at early stages of basin growth.

Map patterns in the southeast of the domain, deduced largely from magnetic data, reveal a progressive increase in strain toward the neighboring Kolwezi domain. An array of ENE-trending décollement folds root into a layer-parallel detachment within the upper Grand Conglomerate (Fig. 13a). Fold profiles diminish in amplitude and increase in wavelength to roughly the latitude of the Kamoa deposit, where the detachment appears to die out, implying a phase of NNW-directed transport that was likely synchronous with amplification of domes and basins in the region of the deposits. Another possible example of décollement folding occurs on the eastern flank of the Kibaran inlier, where NW-trending folds have longer wavelength and are apparently rooted into uppermost levels of the Mwashya Subgroup or lower Grand Conglomerate. In both examples, the detachments occur well above the stratigraphic level of primary salt, but it is suspected that they cut down-section to this level toward the basin’s interior (i.e., central Congolese Copperbelt) in a mirror image to that demonstrated in the southern Congolese Copperbelt. They may reflect nucleation at rheological interfaces or decoupling along small stratabound salt allochthons that have cut upsection from parent salt at the basin’s periphery. Other than the latter scenario, there is no evidence of halokinesis in the Kamoa domain, indicating that primary salt was either absent (i.e., nondeposited) or insufficiently thick to be destabilized during later stages of basin growth or orogenesis.

Kolwezi domain: The Kolwezi domain possesses one of the most distinctive structural features of the Congolese Copperbelt—the presence of extensive, lobate klippen (map extents of <30–700 km2), preserving Roan (including Mines Subgroup) to locally middle Nguba Group strata, which ubiquitously overlie thrust flats positioned ~200 m above the base of the Ngule Subgroup (Desemaecker et al., 1963; A.P. François, unpub. report, 1973; François and Cailteux, 1981; Kampunzu and Cailteux, 1999; Jackson et al., 2003). The configuration thus has similarities to that of the central Congolese Copperbelt; however, the dimensions of the klippen to the south are considerably smaller and in all cases geometrically linked to adjacent diapiric antiforms.

It can be seen from Figure 13a that the distribution and form of klippen are controlled in part by the fold architecture of the neighboring Kamoa domain. The Kolwezi and Tombolo klippen, in particular, occupy the cores of open, doubly plunging synforms developed at the same structural level as the Kamoa domain. In the case of the Kolwezi klippe, the structural depression is controlled partly by the interference of the previously documented ENE- and SE-plunging décollement folds; i.e., the klippe is located at the cross-strike projection of the northern limit of intra-Grand Conglomerate detachment in the Kamoa domain. The klippe’s footwall is conventionally shown on maps and sections to comprise a laterally continuous sheet; however, observations from deep mine levels and sparse drill hole intercepts of Ngule Subgroup strata indicate that steeply bounded breccia bodies and subvertical bedding occur locally, with a structural grain that is sympathetic with the dominant east-northeast synformal axis (comments and discussion by J. Cailteux in François, 1994). On its southern margin, the klippe is bounded by a zone of high-amplitude, upright, strongly noncylindrical folds, again with both ENE- and NW-trending axes, and breccia-cored antiformal culminations (Manika complex; Fig. 13a). Also occupying the same structural level as the Kamoa domain to the west, this complex has clear geometric affinities to the central Congolese Copperbelt and is interpreted as diapiric in origin. This relationship is significant in that it demonstrates that at least the southern part of the Kolwezi domain contained sufficiently thick salt for halokinesis and that a major subbasin boundary must therefore exist between the Manika complex and the Kamoa domain. Further evidence of an eastward increase in accommodation development during Roan Group sedimentation is revealed from deep drilling in the core of the Mamfwe anticline, located east of Kolwezi and within the structural footwall to the klippen (Fig. 13a, b). The closure is an along-axis projection of the NW-trending décollement folds on the eastern margin of the Kamoa domain but here constitutes an antiformal thrust stack that involves folded decoupling surfaces soling into Roan Group evaporites (Cailteux, 1991; Jackson et al., 2003), a facies that is absent to the northwest.

Although each of the klippen preserve Mines Subgroup écailles, only Kolwezi is richly mineralized, accounting for roughly half of the classic Congolese Copperbelt Cu-Co resources (Hitzman et al., 2012). It differs from the rather chaotic form of the central Congolese Copperbelt deposits in that while still complexly deformed, the volume of breccia is low, and a relatively systematic array of inclined to recumbent folds and associated thrusts can be resolved. A map and profile through the klippe are shown in Figure 14, revealing a broad bilateral symmetry to the fold-and-thrust architecture. A downward-converging array of structures project toward the core of the synformal sole thrust, having accommodated apparent northward transport on the northern limb and vice versa on its southern limb. The geometric effect is to superpose a low-amplitude hanging-wall antiformal stack upon an open synformal footwall. There is also a link between metal tenor and this configuration, the highest grade and largest tonnage deposits occurring toward the center of the antiformal stack, diminishing progressively to its peripheries (A.P. François, unpub. report, 1973).

Fig. 14.

Internal structural geometry of the Kolwezi klippe, modified from A.P. François, (unpub. report, 1973). (a) Surface map illustrating preservation of complexly thrusted and folded Mines and lower Dipeta Subgroups strata. The arenaceous stratigraphic footwall to the Mines Subgroup, R.A.T, is unusually coherent and lacks significant volumes of breccia facies. Refer to Digital Appendix Table A2 for lithologic descriptions of units. (b) Klippe profile showing a downward-converging, principally low angle fault array and associated recumbent to inclined folds, producing a low-amplitude antiformal stack centered near the broad synformal closure defined by the form of the sole thrust. Note that some of the faults appear to have accommodated net loss of stratigraphy, a feature interpreted to indicate an extensional history prior to inversion. Highest Cu grades and tonnages occur within the central part of the profile. (c) Restored profile interpreting downward-converging listric normal faults that originally rooted into an attenuated salt layer, now defined by a metric-scale breccia body lining the surface of the sole thrust. The principal resource is positioned at flip in the polarity of the listric faults. The correspondence between metal endowment and structural configuration is considered to in part, at least, reflect fluid focusing through the lower zone of fault convergence. The present configuration records ~50% north-south shortening. Abbreviations: CCCB = central Congolese Copperbelt, C.M.N. = Calcaire à Minerais Noir, Gp = Group, NCCB = northern Congolese Copperbelt, R.A.T. = Roches Argilo-Talqueuses, SCCB = southern Congolese Copperbelt, S.D. = Schistes Dolomitiques, undiff. = undifferentiated, ZCB = Zambian Copperbelt.

Fig. 14.

Internal structural geometry of the Kolwezi klippe, modified from A.P. François, (unpub. report, 1973). (a) Surface map illustrating preservation of complexly thrusted and folded Mines and lower Dipeta Subgroups strata. The arenaceous stratigraphic footwall to the Mines Subgroup, R.A.T, is unusually coherent and lacks significant volumes of breccia facies. Refer to Digital Appendix Table A2 for lithologic descriptions of units. (b) Klippe profile showing a downward-converging, principally low angle fault array and associated recumbent to inclined folds, producing a low-amplitude antiformal stack centered near the broad synformal closure defined by the form of the sole thrust. Note that some of the faults appear to have accommodated net loss of stratigraphy, a feature interpreted to indicate an extensional history prior to inversion. Highest Cu grades and tonnages occur within the central part of the profile. (c) Restored profile interpreting downward-converging listric normal faults that originally rooted into an attenuated salt layer, now defined by a metric-scale breccia body lining the surface of the sole thrust. The principal resource is positioned at flip in the polarity of the listric faults. The correspondence between metal endowment and structural configuration is considered to in part, at least, reflect fluid focusing through the lower zone of fault convergence. The present configuration records ~50% north-south shortening. Abbreviations: CCCB = central Congolese Copperbelt, C.M.N. = Calcaire à Minerais Noir, Gp = Group, NCCB = northern Congolese Copperbelt, R.A.T. = Roches Argilo-Talqueuses, SCCB = southern Congolese Copperbelt, S.D. = Schistes Dolomitiques, undiff. = undifferentiated, ZCB = Zambian Copperbelt.

In detail, it can be seen that some of the low-angle faults appear to have accommodated loss of stratigraphy (Fig. 14b). Although a process of out-of-sequence thrusting could account for this geometry, evidence of subtle facies variation across the faults (A.P. François, unpub. report, 1973) is considered to favor a model of preorogenic extensional collapse, a mechanism reflected in the restored profile (Fig. 14c). Regardless of whether the extensional model is accepted, the lateral continuity of bed forms across thrusts and fold closures demands that the klippe preserves a north-south Roan Group profile that had a precompressional length of >15 km. The 50% internal shortening of the klippe as calculated from the restored profile is in stark contrast to the predominantly gently folded character of the footwall. Any proposed mechanism of thrust sheet emplacement must take into account both the preorogenic lateral extensiveness of the Roan sheet and the manifestly heterogeneous shortening.

Pumpi domain: The eastern boundary of the Kolwezi domain is transitional into the Pumpi domain and marked by an increase in the geometric complexity of thrust slices and preservation of stratigraphy to the uppermost levels of the Katangan basin (Fig. 13). Internally, the domain exhibits a distinctive map pattern involving lozenge-shaped blocks of Nguba and Kundelungu Groups strata enclosed in breccias facies, affected by dominantly WNW- to NNE-trending folds and local thin-skinned thrust repetition. Mines Subgroup écailles are present only in the southern part of the complex to a latitude that corresponds broadly to the northern edge of the Tombolo klippe. Their paucity to the north most likely reflects a lateral facies boundary at Roan Group levels, i.e., nondeposition of typical Mines Subgroup facies associations in the north; however, there is also evidence of a compositional shift in the enclosing breccia facies toward the north, from predominantly Roan Group to Kundelungu Group derivation (A.P. François, unpub. report, 1973; Byrne, 2017). This feature is considered to indicate that the breccia facies preserve progressively higher structural levels of a former salt complex to the north. It can be seen in Figures 13a and b that while the margins of breccia facies locally parallel stratigraphy, most commonly at Grand Conglomerate and Ngule Subgroup levels, they also transgress the intervening stratigraphy at high angles and with curvilinear traces.

The southern margin of the domain partly overthrusts the northern limb of the Mamfwe anticline with layer-parallel contact, which in turn is overthrust along the Kansuki fault zone by the relatively thickened, upright-folded strata of the central Congolese Copperbelt (Fig. 13a, b). Eastward of the Mamfwe anticline, the Kansuki fault zone steepens and ultimately flips polarity, such that fragmental Roan Group strata are exhumed southward over rocks of the central Congolese Copperbelt (Fig. 13a). The northern margin is defined by an abrupt transition to weakly deformed, near-horizontal strata of the upper Kundelungu Group: the Ngule and Bianco Subgroups. Following this boundary from northwest to southeast, an apparently stratabound body of breccia pinches out at the lower level of the Ngule Subgroup. A similar, partly stratabound breccia occurs near the northeastern transition to the Tenke domain. The intervening zone appears to link the relatively undeformed northern Kundelungu Group domain with the complexly folded and thrusted Pumpi domain without obvious structural dislocation.

Immediately inboard of the northern domain boundary is the Mongo structure, a block with similar shape and dimension to the klippen of the neighboring Kolwezi domain but exclusively composed of (at the surface at least) anomalously highly deformed and thick upper Ngule Subgroup strata (A.P. François, unpub. report, 1973). Partitioning of high strains within the Mongo structure, as compared to neighboring regions of similar and lower stratigraphic levels, is demonstrated both in an image of gridded dip magnitudes and in cross section (Fig. 15). The cross section, projected to the interpreted base of the breccia, shows the thickened upper Ngule Subgroup interval occupying an inverted subbasin positioned between lozenges of older Nguba and Kundelungu Groups strata, apparently exhumed toward the south. Successive palinspastic reconstructions to the level of the lowermost Ngule Subgroup reveal the form of a crestal graben, >5 km in width, produced due to the collapse of a diapiric head. The interpretation accounts for the shape of the intervening blocks and the relative positions, geometries, and compositional variations of breccia bodies through the development of a salt stock canopy that had fully enclosed lozenge-shaped rafts at the onset of Ngule Subgroup sedimentation. The rapid northward transition to relatively undeformed, condensed strata is thus interpreted to record yet another major subbasin edge, with limited or no parent salt outside of the Pumpi domain (cf. Jackson et al., 2003).

Fig. 15.

Geometry of the Mongo structure (see also Fig. 13 for location). (a) Gridded bedding dip angles from the northern Pumpi domain, illustrating concentration of rotational strain within the Mongo structure. Neighboring areas of Kundelungu Group strata have contrastingly low bedding dip magnitudes. (b) Palinspastic reconstruction of the Mongo structure, revealing thickened Ngule Subgroup strata within a crestal graben, i.e., collapsed diapiric head. During early Ngule Subgroup deposition, salt encompassed rafts of the Nguba and lower Kundelungu strata; a salt stock canopy was emplaced at the level of the lower Ngule Subgroup. Subsequent diapiric collapse is interpreted to have been driven by extensional separation of the neighboring rafts. Abbreviations: CCCB = central Congolese Copperbelt, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 15.

Geometry of the Mongo structure (see also Fig. 13 for location). (a) Gridded bedding dip angles from the northern Pumpi domain, illustrating concentration of rotational strain within the Mongo structure. Neighboring areas of Kundelungu Group strata have contrastingly low bedding dip magnitudes. (b) Palinspastic reconstruction of the Mongo structure, revealing thickened Ngule Subgroup strata within a crestal graben, i.e., collapsed diapiric head. During early Ngule Subgroup deposition, salt encompassed rafts of the Nguba and lower Kundelungu strata; a salt stock canopy was emplaced at the level of the lower Ngule Subgroup. Subsequent diapiric collapse is interpreted to have been driven by extensional separation of the neighboring rafts. Abbreviations: CCCB = central Congolese Copperbelt, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Tenke domain: Eastward of the Pumpi domain, Roan Group strata are exposed in a broad structural window, host to the giant Tenke and Fungurume deposits, wherein overlying Nguba and Kundelungu Groups have been largely lost to erosion (Schuh et al., 2012; Figs. 13a, 16a). This relatively deep level of exposure into the structural pile relates in part to yet another change in structural style, dominated by highamplitude, low-wavelength upright folding. While considerably more complex in the Roan Group écailles, this lateral transition from thin-skinned to thick-skinned deformation is also demonstrated to the east of the Tenke domain, where the window passes into a zone preserving higher structural and stratigraphic levels involving high-wavelength, upright synforms, cored by Kundelungu Group strata.

Fig. 16.

Balanced cross section of the northern and central Congolese Copperbelt through the Tenke domain, looking west. Refer to Figure 13 for the section trace. (a) The present configuration involves ~35% orogenic shortening (or a stretch of 0.65), with <4-km displacement on the northern thrust front. Major deposits are projected onto section. Those positioned south of the Kansuki fault zone occupy the cores of the pinched diapirs. The Tenke deposits, exposed in a tectonic window, are shown to occur below a carapace of relatively condensed Nguba and Kundelungu Groups strata, now largely lost to erosion. (b) The restored profile reveals a southward-thickening series of minibasins, bounded above the Lower Roan Subgroup synrift package by an array of high-amplitude salt walls (i.e., Mines and Dipeta Subgroups and salt-related breccia). Minibasins are interpreted to have welded at the bases of their thickened cores to the synrift package. Hosts to classic Cu-Co ores are considered to have originated close to salt welds and were repositioned to higher structural levels via salt flow and/or orogenic thickening of the pile. (c) Expanded views of the Tenke domain with schematic representation of Mines Subgroup écailles. The restored profile shows the Kansuki fault zone as a diapiric structure positioned above a major synrift subbasin boundary and between basal-welded minibasins. The northern minibasin, about which known high-grade Cu mineralization occurs, corresponds to the northern limit of Mines Subgroup écailles (see also Fig. 13). The present configuration shows inflation and internal buckling of the Tenke upper Roan Group sequence due to northward impingement of the southern minibasin and partial underthrusting of the Kansuki fault zone. A Grand Conglomerate-cored syncline is the only preserved remnant of the Nguba-Kundelungu Groups carapace in the region of the Tenke deposits and is interpreted to represent the keel of an inverted minibasin. Abbreviations: CCB = Congolese Copperbelt, FZ = fault zone, Gp = Group, Kund. = Kundelungu, PGE = platinum group element, Sgp(s) = Subgroup(s).

Fig. 16.

Balanced cross section of the northern and central Congolese Copperbelt through the Tenke domain, looking west. Refer to Figure 13 for the section trace. (a) The present configuration involves ~35% orogenic shortening (or a stretch of 0.65), with <4-km displacement on the northern thrust front. Major deposits are projected onto section. Those positioned south of the Kansuki fault zone occupy the cores of the pinched diapirs. The Tenke deposits, exposed in a tectonic window, are shown to occur below a carapace of relatively condensed Nguba and Kundelungu Groups strata, now largely lost to erosion. (b) The restored profile reveals a southward-thickening series of minibasins, bounded above the Lower Roan Subgroup synrift package by an array of high-amplitude salt walls (i.e., Mines and Dipeta Subgroups and salt-related breccia). Minibasins are interpreted to have welded at the bases of their thickened cores to the synrift package. Hosts to classic Cu-Co ores are considered to have originated close to salt welds and were repositioned to higher structural levels via salt flow and/or orogenic thickening of the pile. (c) Expanded views of the Tenke domain with schematic representation of Mines Subgroup écailles. The restored profile shows the Kansuki fault zone as a diapiric structure positioned above a major synrift subbasin boundary and between basal-welded minibasins. The northern minibasin, about which known high-grade Cu mineralization occurs, corresponds to the northern limit of Mines Subgroup écailles (see also Fig. 13). The present configuration shows inflation and internal buckling of the Tenke upper Roan Group sequence due to northward impingement of the southern minibasin and partial underthrusting of the Kansuki fault zone. A Grand Conglomerate-cored syncline is the only preserved remnant of the Nguba-Kundelungu Groups carapace in the region of the Tenke deposits and is interpreted to represent the keel of an inverted minibasin. Abbreviations: CCB = Congolese Copperbelt, FZ = fault zone, Gp = Group, Kund. = Kundelungu, PGE = platinum group element, Sgp(s) = Subgroup(s).

The northern margin of the window is an ENE-striking imbricate thrust stack, involving a relatively condensed sequence of Nguba and Kundelungu Groups strata (Figs. 13a, 16a). As is the case with the northern fringe of the Pumpi domain, strain dissipates rapidly northward of the thrust front, which at its maximum is calculated to have accommodated ~4 km of NNW-directed transport. The southern margin is more complex, defined in part by the Kansuki fault zone and along strike to the east by an ENE-striking fault that transects tightly folded Nguba and Kundelungu Groups strata at high angle. The boundary between these two fault segments is the Kakanda-Luisha fault zone, which projects northwestward through the Tenke domain (evidenced by the local northwest structural grain of écailles) to the northeastern margin of the Pumpi domain (see also Fig. 8b).

The boundary with the Pumpi domain also includes discontinuous fault segments with northwest strike but is principally defined by the progressive eastward amplification of the Roan Group-cored window. Remnant lower Nguba Group strata are preserved in the cores of isolated, doubly plunging, upright, NE- to ENE-trending synforms in the western and central parts of the domain (Figs. 13a, 16a, c). The trace of these synforms corresponds to the northern limit of Mines Subgroup écailles and, where intersecting with the northwestern projection of the Kakanda-Luisha fault zone, localizes the largest of the Tenke group of deposits currently in production. Partly brecciated strata located in the northern part of the domain are correlated with the reductant-poor Dipeta Subgroup, raising the potential that Nguba Group-cored synclines mark a change in structural level, with fertile Mines Subgroup strata still preserved at depth (Schuh et al., 2012). While this scenario cannot be discounted, the apparent continuity of the northern trace of Mines Subgroup écailles from Fungurume to Tombolo, a distance of >100 km across domains of highly contrasting structural style and levels of exhumation, makes it unlikely. An interpretation that the synclines correspond with a fundamental facies change in the Roan Group is preferred; the position and morphology of the synclines are thus considered to have been inherited from deeper-seated structures that controlled growth during Roan Group sedimentation.

The structural morphology of Mines Subgroup écailles shares similarities with that of the Kolwezi klippe in that it involves relatively minor volumes of breccia and a systematic, albeit complex, arrangement of thrusts and folds (François, 1986; Schuh et al., 2012). Moreover, the longitudinal strain affecting the écailles is considerably greater than that recorded by the Nguba-Kundelungu Groups carapace preserved within and at the margins of the domain. The principal difference with Kolwezi is the form and history of folding, which include upright box-shaped profiles, amplified to >2 km, that overprint a prior phase of layer-subparallel thrusting. The most complex patterns affect the high-grade Tenke deposit cluster (Schuh et al., 2012) at the margins of Nguba Group-cored synclines. Here, thrusts, which duplicate the Roan Group stratigraphy along a decoupling surface at the base of the Mines Subgroup, are folded about isoclinal closures (Fig. 17). At high structural levels, axial surfaces are shallowly inclined to recumbent, rotating at depth to subvertical. The higher-level, shallowly inclined parts of folds in places display opposing vergences, imparting mushroom-shaped profiles (Schuh et al., 2012). Where antiformal, these profiles superficially resemble diapiric structures; however, that they manifestly postdate thrusting and are cored by thrust-duplicated stratigraphy requires nucleation via shortening rather than the relative rise of salt. Furthermore, while shortening of the écailles is greater than that of Nguba Group carapace, the orientations of macroscopic fold axial traces are similar, suggesting that contrasting deformation styles are not completely decoupled.

Fig. 17.

Cross section of the Kwatebala deposit, easternmost of the Tenke deposits (refer to Fig. 13a for deposit location; modified from François, 1986). The geometry involves an early stage thrust stack, refolded by tight upright closures and open recumbent closures. Ore occurs within the Kamoto dolomite and lowest levels of the S.D.-C.M.N. sequence, and in this case, only the structurally highest thrust sheet contains economic mineralization. It is interpreted that part of the early thrust history relates to localized shortening at the margins of downward-impinging, salt withdrawal basins. Abbreviations: C.M.N. = Calcaire à Minerais Noir, S.D. = Schistes Dolomitiques, Sgp = Subgroup.

Fig. 17.

Cross section of the Kwatebala deposit, easternmost of the Tenke deposits (refer to Fig. 13a for deposit location; modified from François, 1986). The geometry involves an early stage thrust stack, refolded by tight upright closures and open recumbent closures. Ore occurs within the Kamoto dolomite and lowest levels of the S.D.-C.M.N. sequence, and in this case, only the structurally highest thrust sheet contains economic mineralization. It is interpreted that part of the early thrust history relates to localized shortening at the margins of downward-impinging, salt withdrawal basins. Abbreviations: C.M.N. = Calcaire à Minerais Noir, S.D. = Schistes Dolomitiques, Sgp = Subgroup.

As is the case for Kolwezi, the pronounced strain partitioning between Roan Group and enclosing strata is problematic. In order to preserve bed lengths in reconstruction of the structural profile, the intensely shortened Roan Group écailles must be balanced laterally by considerably lower compressional strains or possibly even extension. The mechanism of lateral salt withdrawal to generate welds, or inversion of resultant geometries, provides a possible solution. The restored cross sections in Figures 16b and c interpret the array of Nguba Group-cored synclines as the erosional remnants of minibasins, which through a process of differential loading led ultimately to welding with underlying Roan synrift strata; i.e., salt-bearing strata, including the Mines Subgroup, were laterally displaced toward the lower flanks of intervening diapirs. In this scenario, it is permissible that shortening of écailles initiated during extensional phases of basin growth and the removal of material below the minibasins balanced by low-magnitude thrusting and buckling at their peripheries. Continued strain partitioning is likely to have occurred during basin closure, with lateral translation of welded minibasins atop the synrift package bulldozing and refolding écailles within pinched and amplified diapirs (Fig. 16c). The overall effect of concentrating shortening strains between weld points is shown to jack up the Nguba-Kundelungu Groups carapace, potentially accounting for the relatively deep structural level exposed throughout the Tenke domain.

High-magnitude thrust transport or basin inheritance?

Preservation of areally extensive klippen and compelling evidence for extensive thrust flats in the Kolwezi domain have provided the basis for conventional models of thin-skinned, largely N-directed thrust emplacement throughout the Congolese Copperbelt (Desemaecker et al., 1963; A.P. François, unpub. report, 1973; François and Cailteux, 1981; Kampunzu and Cailteux, 1999; Porada and Berhorst, 2000; Jackson et al., 2003; Schuh et al., 2012). The magnitude of displacement varies between the models (~40–150 km), but essentially each interprets the expanses of Roan Group strata and their Nguba-Kundelungu carapaces exposed throughout the Tenke, Pumpi, and Kolwezi domains (and currently exposed levels southward of the Kansuki fault zone) as allochthonous relative to the Kibaran inlier and its unconformably overlying Katangan succession (i.e., Kamoa domain). The models are problematic in that while the geometries of the Kolwezi domain demand a thin-skinned thrust architecture, there is little evidence of a high-magnitude thrust front farther east; <4-km N-directed transport was accommodated on the northern margin of the Tenke domain, whereas parts of the northern Pumpi domain show no evidence of thrust dislocation. This paradoxical relationship has been unconvincingly explained by synorogenic deposition of molasse-type strata on an emergent thrust front (e.g., A.P. François, unpub. report, 1973; Jackson et al., 2003).

With the discovery of the Kamoa deposit in 2008 (Broughton and Rogers, 2010), the allochthonous models became seriously questioned, if not untenable. Direct spatial and geometric links between Kamoa and Kolwezi (i.e., <10-km, along-structural-grain separation of Kamoa from the westernmost Kolwezi deposits) make it highly unlikely that the systems were originally separated by >40 km, unless mineralization occurred late in the orogenic history and was superimposed upon a preexisting thrust stack. Apparent projection of major lineaments, including the Monwezi and Kakanda-Luisha fault zones, across the Kansuki fault zone to intersect world-class deposits in the northern Congolese Copperbelt (Kakula and Tenke, respectively) also militates against high-magnitude thrust transport. Indeed, the systematic relationship between isopach gradients and these structures strongly implies that they represent basin-bounding elements within a setting that has preserved its original and fundamental extensional form, with major deposits contained in its condensed peripheries (Figs. 12a, 16b).

The analysis above of particular structural elements from the Tenke and Pumpi domains provides potential clues as to the emplacement mechanism of thrust sheets in the Kolwezi domain. The salt stock canopy configuration, interpreted to account for the style of the Pumpi domain, invokes the development of a regional, stratabound zone of weakness at the top of the stratigraphic pile prior to significant shortening (Fig. 15b). Punctuated by stratigraphically transgressive salt walls, this upper zone of potential dislocation would have been linked to the parent salt layer at depth across a 3-D network of steeply dipping ramps with duplex configuration. With the onset of significant shortening, this framework would have led to vertical exhumation ± relatively minor lateral translation of intervening rafts, as manifest by the stacked thrust arrangement of the Pumpi domain (Figs. 13b, 15b).

There are elements of this configuration in the central Congolese Copperbelt that provide evidence that progressively evolving diapir heads migrated and/or coalesced laterally during lower Ngule Subgroup sedimentation to produce stratabound breccia complexes that consist predominantly of either Roan or Kundelungu Groups fragments. The principal difference is the respective thick- versus thin-skinned character of shortening-related strains in the central and northern Congolese Copperbelt. This feature likely relates to the contrasting thickness of suprasalt stratigraphy in the two basin compartments (e.g., Figs. 12a, 16a, b). Thin suprasalt stratigraphy and consequential low-amplitude diapirs to the north allowed for transport on short-length ramps to duplicate the pile, whereas lateral impingement of thick suprasalt intervals across pinched diapirs with lateral welds (e.g., Jackson et al., 2003) in the southern depocenter maximum promoted upright folding.

The relatively thin character of the suprasalt sequence in the northern Congolese Copperbelt was likely matched by thin parent salt. The apparent late-stage diapiric collapse of the Mongo structure is considered to indicate that, at least immediately prior to complete basin inversion, there was insufficient salt in the parent layer to maintain the diapiric head. The palinspastic model in Figure 15b invokes a final stage (i.e., upper Ngule Subgroup sedimentation) of extension and separation of Nguba-Kundelungu Groups rafts in having limited the supply of salt necessary to prevent diapir collapse (e.g., Vendeville and Jackson, 1992); however, dissolution of a broad, emergent diapiric structure could have produced a similar configuration, without the interpreted extensional phase.

The lateral transition from the Pumpi domain to the Kolwezi domain involves a significant increase in the areal extent of the remnant thrust sheets and an apparently simpler, less fragmental, fold-dominated architecture of the lower sheet. Applying the Pumpi domain structural model, this change in structural style could be explained by an increase in the spacing of salt walls. This, in turn, is potentially related to thinner parent salt, a concept consistent with evidence of a westward pinch-out of salt-bearing stratigraphy into the Kamoa domain. Taking into account these constraints and the previously described structural peculiarities of the Kolwezi klippe, a model for the thrust sheet emplacement is shown in Figure 18. It invokes essentially vertical emplacement of a raft and its Mines Subgroup substrate atop a symmetrically downward ramping fault system, the latter potentially having accommodated syn-Ngule Subgroup extensional collapse. The model accounts for the bilateral symmetry of the klippe’s fold-and-thrust array and the unusual juxtaposition of an antiformal stack above a broadly synformal footwall. The latter is interpreted as an amalgam of rafts, the synformal axis nucleated along a vertical salt weld. Although drill hole density at this level is presently insufficient to validate this footwall configuration, the documented evidence of transgressive breccia bodies and domains of anomalously high bedding dip at the base of the klippe (e.g., comments and discussion by J. Cailteux in François, 1994) make it permissible.

Fig. 18.

Schematic model for the emplacement of the Kolwezi klippe (looking west): present configuration, and profile restored to the level of the upper Ngule Subgroup. Approximate position of section indicated in Figure 13. The presently preserved antiformal stack of Mines Subgroup écailles is interpreted to have been positioned at the base of rafted minibasin, the latter now lost to erosion. The highest-grade core of the klippe is shown to have been symmetrically arranged about a salt weld prior to shortening. During the final shortening phase, bounding rafts converged to weld laterally, causing the klippe to be emplaced vertically to the level of an Ngule Subgroup salt stock canopy. Abbreviation: Sgp = Subgroup.

Fig. 18.

Schematic model for the emplacement of the Kolwezi klippe (looking west): present configuration, and profile restored to the level of the upper Ngule Subgroup. Approximate position of section indicated in Figure 13. The presently preserved antiformal stack of Mines Subgroup écailles is interpreted to have been positioned at the base of rafted minibasin, the latter now lost to erosion. The highest-grade core of the klippe is shown to have been symmetrically arranged about a salt weld prior to shortening. During the final shortening phase, bounding rafts converged to weld laterally, causing the klippe to be emplaced vertically to the level of an Ngule Subgroup salt stock canopy. Abbreviation: Sgp = Subgroup.

Although controlled by diapiric architecture, the emplacement mechanism is not halokinetic. In fact, the preinversion position of the core of the écaille complex is placed at the welded base of the raft (i.e., an interdiapiric position; Fig. 18), with a structural style and high-grade ore distribution akin to that interpreted for the Tenke group of deposits.

The model also requires that the traces of vertical welds be mappable throughout the Kolwezi domain, in particular projecting outward from, or at least in the vicinity of, the klippen. One such zone potentially exists ~2 km south of the Tombolo klippe, defined by an ENE-trending alignment of narrow breccia bodies enclosed by Ngule Subgroup strata (Fig. 13a). The apparent paucity of similar structures may be attributed to poor outcrop and the typically recessive character of breccia facies, but these structures are expected to conform principally to synformal closures and localized abrupt stratigraphic offsets and to define a polygonal fracture array. The pattern of ENE- and NW-trending folds that enclose the Kolwezi klippe (Fig. 13a) is potentially inherited from such an array.

Macroscopic Basin Configuration and the Spatial Distribution of Ore

While outstanding issues remain in the validation of the structural models, the weight of evidence favors a parautochthonous configuration for at least the Congolese Copperbelt. Perhaps most compelling in this regard is the lateral continuity and lack of mutual offset affecting major lineaments. The systematic relationship between isopach gradients and these structures strongly implies that they represent basin-bounding elements within a setting that has largely preserved its extensional form. Fundamental elements of extensional architecture in the central and northern Congolese Copperbelt include (1) a progressively northward-tapering stratigraphic profile, (2) a halokinetic framework affecting middle and upper levels of the basin fill, southward of a salt-poor, condensed periphery, and (3) evidence of salt welding at the base of down-building minibasins in the form of heterogeneous strain partitioning in salt-bearing strata (Fig. 19a).

The southern margin of the Congolese Copperbelt depocenter is less well constrained in terms of its extensional geometry, due in part to poorer exposure and less detailed mapping. Décollement folds on the northern flank of the Luina dome have broad affinities with diapiric structures of the central Congolese Copperbelt; however, there are presently insufficient stratal thickness data to demonstrate a relationship between syndepositional diapirism and fold closures. Nor are the data sufficient to reveal the presence of major subbasin-bounding faults; however, the southward disappearance of Mines Subgroup strata, similar to that observed on the northern fringe of the basin, coupled with paucity of halokinetic geometries in the adjacent Zambian Copperbelt where the Katangan sequence is manifestly thin, points toward the presence of one or a series of basin-bounding elements in the region of the Luina dome.

The northern margin of the Domes region is also problematic. Significant northward thickening of the Lower Roan synrift package, demonstrable between the basement domes and Kansanshi, and the appearance of volumetrically significant breccia bodies immediately north of the latter are consistent with a transition into a thick, salt-bearing Congolese basin; however, the high metamorphic grade and evidence of major thrust and nappe repetition permit a significant volume of the depocenter to have been have been exhumed southward and lost to erosion. The abrupt N-dipping basin margin structure depicted in Figure 19a and the overall preservation of the basin’s integrity throughout orogenesis in this region should thus be treated with caution. In essence, however, the gross architecture of a symmetrical basin form, with progressively limited accommodation development and originally thinner salt toward northern and southern peripheries, is considered robust, incorporating all presently available stratigraphic and geometric constraints.

Figure 19b through d presents an attempt to reconcile the various structural styles developed during orogenesis. In the southern parts of the basin, the initial stages of burial and basement-involved thrust duplication recorded in the Domes region (~535 Ma) are equated to prefoliation layer-subparallel decoupling and localized generation of recumbent folds in the Zambian Copperbelt (Fig. 19b). Removal of stratigraphy along the ramp on the northwestern flank of the Zambian Copperbelt is interpreted to indicate relatively low magnitude lateral exhumation of the neighboring depocenter maximum, where the decoupling surface stepped down-section to a stratigraphically lower evaporitic horizon at the top of the Lower Roan Subgroup synrift package.

Drawing a temporal link to the strain history of the Congolese Copperbelt is more equivocal, given that halokinetic deformation commenced at least as early as the second rift climax at ~765 to 740 Ma. It is probable, however, that diapirs were squeezed, amplified, and laterally welded during initial stages of basin closure (e.g., Jackson et al., 2003). Widespread emplacement of breccia facies at a common stratigraphic level (i.e., lower Ngule Subgroup), supports the Jackson et al. (2003) model of a time-specific, likely orogenically controlled, change in diapir configuration from passive vertical amplification to high-level lateral growth (Fig. 19b). Evidence of a parautochthonous orogen favors the model of salt stock canopy emplacement via lateral amalgamation of flattened diapiric heads over lateral extrusion of a far-traveled salt allochthon (i.e., >65 km; Jackson et al., 2003). Small-scale lateral salt allochthon emplacement is potentially recorded at the northern periphery parent salt (Fig. 19b), where layer-parallel decoupling occurs in nonevaporitic strata of the upper Grand Conglomerate. While it is possible that some particularly high amplitude diapirs became emergent prior to orogenesis, the regional stratabound character of the canopy suggests that widespread emplacement of salt at the Earth’s surface was likely during Ngule Subgroup deposition.

The subsequent diapir collapse stage potentially records dissolution of salt deep within the profile. While this process alone had the capacity to generate upper Ngule Subgroup-filled crestal grabens, a component of extensionally driven collapse is favored (Fig. 19c). The principal reason is that lateral separation of intervening rafts (as opposed to welding) would have facilitated differential raft exhumation and emplacement on thrust flats during subsequent shortening. Accepting this reasoning, the extensional phase provides a possible tectonostratigraphic record of decompression, unroofing, and top-block-north sense of shear in the Domes region.

The final stages of basin closure (Fig. 19d) are interpreted to be recorded throughout the Katangan basin as (1) high-wavelength, basement-involved upright folding and associated foliation development in the Zambian Copperbelt, (2) recumbent to upright folding and retrogressive fabric development in the Domes region, (3) tightening of synformal minibasins and likely continued emplacement of high-level thrusts at Ngule Subgroup levels in the central Congolese Copperbelt, (4) folding of, and local thrust duplication along, the salt stock canopy, leading to raft exhumation in the salt-bearing subbasins of the northern Congolese Copperbelt, and (5) subbasin inversion on the salt-poor northern basin fringe, leading to noncylindrical gentle to open upright folds and associated upright foliation development and localized N-directed décollement folding. The model presented thus equates much of the shortening-related strain to postpeak metamorphic stages of the Domes region (i.e., <535 Ma).

Fig. 19.

Schematic model for the evolution of the Katangan basin through orogenesis. The profiles look broadly westward. (a) Basin form immediately prior to orogenesis, involving crude symmetrical tapering about a central depocenter maximum. Diapirs punctuate upper levels of the basin fill in the Congolese Copperbelt, reaching maximum amplitude in central and southern parts. Thin salt in the Domes region and Zambian Copperbelt remained essentially unmodified. The northern basin margin, which records onset of growth during Mwashya Subgroup deposition (marginal coarse-grained siliciclastics), was devoid of salt. (b) D1 shortening, involving burial and high-grade metamorphism of the Domes region, and basin-wide decoupling at the level of salt. Diapirs were squeezed in the Congolese Copperbelt to become fully emergent, linking at the surface in northern parts to produce a salt stock canopy at lower levels of the Ngule Subgroup. Emplacement of a salt allochthon at the level of the Grand Conglomerate at the northern edge of the basin. (c) Extensionally driven decompression in the Domes region and extensional collapse along the evaporitic décollement to produce crestal grabens in the northern Congolese Copperbelt. (d) D2 shortening, involving basin-wide foliation development. Thick-skinned, basement-involved folding of the Zambian Copperbelt, low-angle thrusting in the Domes region partly associated with southward exhumation of the depocenter maximum, continued emergence of diapirs, and vertical stacking of rafts in the northern Congolese Copperbelt. Distribution of deposit types shown with no temporal significance implied. See text for discussion. Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 19.

Schematic model for the evolution of the Katangan basin through orogenesis. The profiles look broadly westward. (a) Basin form immediately prior to orogenesis, involving crude symmetrical tapering about a central depocenter maximum. Diapirs punctuate upper levels of the basin fill in the Congolese Copperbelt, reaching maximum amplitude in central and southern parts. Thin salt in the Domes region and Zambian Copperbelt remained essentially unmodified. The northern basin margin, which records onset of growth during Mwashya Subgroup deposition (marginal coarse-grained siliciclastics), was devoid of salt. (b) D1 shortening, involving burial and high-grade metamorphism of the Domes region, and basin-wide decoupling at the level of salt. Diapirs were squeezed in the Congolese Copperbelt to become fully emergent, linking at the surface in northern parts to produce a salt stock canopy at lower levels of the Ngule Subgroup. Emplacement of a salt allochthon at the level of the Grand Conglomerate at the northern edge of the basin. (c) Extensionally driven decompression in the Domes region and extensional collapse along the evaporitic décollement to produce crestal grabens in the northern Congolese Copperbelt. (d) D2 shortening, involving basin-wide foliation development. Thick-skinned, basement-involved folding of the Zambian Copperbelt, low-angle thrusting in the Domes region partly associated with southward exhumation of the depocenter maximum, continued emergence of diapirs, and vertical stacking of rafts in the northern Congolese Copperbelt. Distribution of deposit types shown with no temporal significance implied. See text for discussion. Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

The spatial distribution of ores within the inverted basin framework is illustrated in Figure 19d (note that no temporal significance is implied). Giant and supergiant classic stratiform Cu-Co ores of the Zambian and Congolese Copperbelts are partitioned into the margins of the depocenter maximum. Zambian Copperbelt ores occupy a condensed basin-margin position, below the level of relatively thin parent salt, and are located in small reducing subbasins or physical hydrocarbon traps, in both cases enclosed in relatively coarse grained subarkosic strata (Selley et al., 2005). The largest of the intrasalt Congolese Copperbelt ores occur close to the lateral termination of thick parent evaporites, whereas smaller, more fragmental varieties are situated in diapiric crests positioned above major subbasin-bounding fault zones. This northward trend toward larger systems is potentially a function of structural style and level of exposure and not a reflection of original dimensions or yet-to-be-identified resources at depth. The considerably thicker suprasalt sequence southward of the Kansuki fault zone and thick-skinned structural style mean that the original stratigraphic level of ore is never exposed. Deposits in this region are restricted to corridors wherein diapirism has effectively lifted the fragmented components of presumably more extensive ores to higher level in the crust. Within the central depocenter maximum, the belt of U-Ni-Co (Au-PGE) ore (e.g., Shinkolobwe) is also intimately associated with high-amplitude diapirs (Fig. 19d).

Outboard of the classic ores, stratiform Cu mineralization steps symmetrically to the base of the Grand Conglomerate, the stratigraphically lowest reducing element within these salt-poor peripheral subbasins that nucleated during the second rift climax (Fig. 19d). In the case of the supergiant Kamoa system, the known deposits occur at or southward of the northern edge of hanging-wall décollement folds, raising the possibility that they were originally overlain by a carapace of allochthonous salt. At the base of the ore, the host-reducing element (basal Grand Conglomerate) is directly underlain by Mwashya Subgroup subarkoses, which are shown schematically in Figure 19d to link into the Lower Roan synrift package toward the basin’s interior. Relatively meager mineralization at Fishtie lies directly on basement (Hendrickson et al., 2015), implying that hydrologic connectivity with the depocenter maximum through a subore aquifer was an important factor in controlling the scale of the ore.

Shear-zone-hosted Cu-U ores at Lumwana occupy the core of a dome, composing structurally interleaved basement and Katangan strata (Fig. 19d). The host is ubiquitously basement gneiss, although in the highest-grade deposit, Malundwe, the orebody is in direct tectonized contact with Lower Roan synrift strata. This relationship again raises the possibility that connectivity with an aquifer deep within the basin profile played a role in ore formation.

Predominantly vein- and/or fracture-hosted mineralization types are widespread across the basin profile but are largely restricted to stratigraphic positions above the level of parent salt. Congolese examples possess a ubiquitous spatial association with breccia facies and include Cu-Co ores hosted primarily within Mwashya Subgroup strata (e.g., Tilwezembe) and lower Nguba Group-hosted Pb-Zn-Cu-Ge ore (e.g., Kipushi), positioned toward the margin and interior of the depocenter maximum, respectively (Fig. 19d). The vein-hosted Cu-(Mo-Au) mineralization type is known only from the Domes region and Zambian Copperbelt (Fig. 19d), where it is widespread but rarely ore forming. Two ore systems of this type, Kansanshi and Frontier, are hosted in lower Nguba Group and Mwashya Group strata, respectively.

In consideration of the gross basinal and stratigraphic partitioning of mineralization relative to salt, ore-forming systems tend to occur in five general locations: (1) extreme basin margin position, outboard of primary salt (Grand Conglomerate-hosted stratiform Cu), (2) basin margin position, below primary salt (classic stratiform Zambian Copperbelt Cu-Co; Roan Group-hosted Cu and/or Ni and basement-hosted Cu-U), (3) basinal margin (possibly transitioning to basin interior) position, enclosed within relatively thin salt (classic stratiform Congolese Copperbelt Cu-Co), (4) basin interior position, intimately associated with diapiric breccias sourced from thick parent salt (Roan Group-hosted U-Ni-Co [Au-PGE]; basal Nguba Group-hosted Pb-Zn [Cu-Ge]), and (5) basin margin position, above the level of primary salt (Mwashya Supergroup/basal Nguba Group-hosted vein-controlled Cu [Mo-Au] and Cu-Co). With the exception of extreme basin marginal ores, the systems occur within a ~500-m vertical distance from the position of either parent or remobilized salt. The following sections examine the hydrologic role of salt in ore formation in relative time and space.

Stratigraphic Partitioning of Fluids and Sources of Salinity

The macroscopic partitioning of ore types is matched by fundamental differences in alteration style, both in stratigraphic sense and laterally throughout the basin. Metasomatism is particularly widespread and intense in association with the stratiform ore types, positioned in subsalt, intrasalt, and saltperipheral parts of the basin. Broadly stratabound in character, the alteration zones extend hundreds of meters into stratigraphic footwalls to ore, over lesser widths into their hanging walls, and on a kilometric scale laterally. The predominantly fracture controlled ore types are typically associated with more localized, smaller-scale alteration patterns, which are most commonly partitioned within or above the highest level of primary salt but at least locally occur in subsalt positions.

The chemical and mineralogic signatures of these contrasting alteration styles are well documented and are almost universally attributed to the passage of evaporitic brines (Darnley, 1960; Bartholome et al., 1973; Cailteux, 1977; Moine et al., 1986; Selley et al., 2005; El Desouky et al., 2009; Hitzman et al., 2012; Capistrant et al., 2015; Halley et al., 2016). In this section, we address the scale of alteration (based on previous work) within the context of the proposed structural model and present new fluid inclusion data from a variety of ore types in an attempt to constrain the source(s) of salinity.

Stratabound alteration types

The stratabound alteration type is the product of Mg and/or K metasomatism. The relative intensity of these chemical components varies laterally across the profile of the Katangan basin. At the basin margins, subsalt Zambian Copperbelt Cu-Co ores and salt-peripheral Cu ores (e.g., Kamoa) are characterized by intense K metasomatism and subordinate Mg enrichment, reflected mineralogically by a K-feldspar ± muscovite ± phlogopite ± Mg chlorite ± dolomite assemblage (Moine et al., 1986; Selley et al., 2005; Schmandt et al., 2013). In the Zambian Copperbelt, this style of alteration occurs below the zone of decoupling and is intensely developed throughout Lower Roan Subgroup strata but locally affects both basement and Upper Roan Subgroup rocks. At Kamoa, authigenic potassic phases are concentrated at and below the level of ore, diminishing in volume up-section through the Grand Conglomerate.

West of the Zambian Copperbelt in the Domes region, Mg metasomatism dominates in sequences below the level of former salt. Magnesium-bearing minerals, such as talc ± phlogopite ± tremolite ± dolomite ± magnesite, are developed throughout the Lower Roan Subgroup strata. Magnesium metasomatism is most intense in the uppermost part of the basal synrift package, where it occurs in association with hematitic ironstones (Cosi et al., 1992; Steven and Armstrong, 2003; John et al., 2004; Capistrant et al., 2015). The form and scale of the metasomatized compartment were demonstrated by Halley et al. (2016) through analysis of multielement geo-chemical data from a soil survey of the Sentinel-Enterprise district. Positioned below the level of decoupling near the base of the Mwashya Subgroup, the zone of intense Mg metasomatism projects to the surface within an area of >600 km2 and is accompanied by enrichment of rare earth elements and Ni. With the exception of reducing ore-bearing intervals (i.e., Sentinel and Enterprise), Cu values were shown to be strongly depleted relative to calculated detrital concentrations, implying the passage of an oxidizing brine capable of efficiently stripping and transporting metal.

While it is tempting to draw a direct link between regional Mg metasomatism, Cu depletion, and ore formation, the host strata to subsalt ores of the Domes region (both stratiform and shear-hosted) appear to record a less intense Mg signature compared to their immediately bounding units. At Malundwe and Sentinel, the ores are associated with a dominant assemblage of phlogopite-kyanite-muscovite-quartz (Steven and Armstrong, 2003; Hitzman et al., 2012; Bernau et al., 2013), implying either minor addition of K or preservation of the K content of the protoliths. Although further work is required to resolve this paradoxical relationship, we postulate that unusually acidic conditions at the sites of sulfide precipitation may have inhibited the stabilization of Mg-bearing phases.

Basin-marginal, intrasalt classic Congolese Copperbelt Cu-Co ores show a similar partitioning of Mg and K phases. The arenaceous stratigraphic footwalls to ore (maximum preserved thicknesses ≤300 m) are characterized by a relatively simple assemblage of quartz + Mg chlorite + dolomite (Bartholome et al., 1973; Cailteux, 1977; Schuh et al., 2012; Byrne, 2017), which reflects intense Mg metasomatism with the near-complete removal of alkali elements (e.g., Moine et al., 1986). Magnesium alteration transgressed the lower carbonate-dominant ore-bearing interval, now represented by magnesite + dolomite + Mg chlorite, and transitions up-section into a stratabound zone of thorough silicification and textural destruction. Although the lower Mg alteration remains intense, there is an abrupt appearance of low-level alkali concentrations, typically slightly lower than those of primary muscovitic detrital compositions. Through upper argillite-hosted ores and into their assay hanging walls, there is a trend toward progressively lower Mg concentrations, coupled with phengitic mica compositions and, in rare instances, authigenic K-feldspar.

The strong stratigraphic partitioning of alteration zones associated with the classic Congolese Copperbelt Cu-Co ores is interpreted to reflect variations in primary and secondary permeability during progressive up-section infiltration of an Mg-saturated brine. The pattern of intense alteration within and below the level of ore, diminishing into the stratigraphic hanging wall, thus has affinities with that of Kamoa. Importantly, the pattern is preserved in all deposits, regardless of the degree of structural complexity; i.e., Mg metasomatism clearly occurred prior to significant deformation (fragmentation, faulting, folding) of the host sequence and occurred on a lateral scale equivalent to, or greater than, that in the Domes region (i.e., >600 km2). However, breccia facies that partly enclose Mines Subgroup écailles have a matrix of Mg chlorite + sparry dolomite + hematite + quartz (Cailteux et al., 2005; Byrne, 2017), indicating that Mg metasomatism either persisted or recommenced during fragmentation of the intrasalt stratigraphy. The contrasting stratigraphic and transgressive alteration patterns suggest a fundamental change in the macroscopic permeability framework, from primary lithofacies to fracture control, and support arguments for multiphase or protracted fluid input (e.g., Dewaele et al., 2006; El Desouky et al., 2009; Schuh et al., 2012).

In the depocenter maximum, stratigraphic partitioning of alteration at the level of the Mines Subgroup is not apparent. At Shinkolobwe, the entire sequence is affected by unusually intense Mg metasomatism, with magnesite ± Mg chlorite replacing primary phases and forming vein arrays and masses that extend well into the stratigraphic hanging wall of the ores (Derriks and Vaes, 1956; Ball, 2015). The apparent disparity with basin-marginal systems may reflect higher fluid fluxes in the depocenter maximum, such that stratigraphically partitioned alteration patterns either never formed or were destroyed during the ongoing alteration that accompanied fragmentation.

Structurally controlled alteration types

Sodic metasomatism is widespread at sub-middle Nguba Group levels but is intensely developed only in the Domes region and Zambian Copperbelt. In the latter, it is recorded by albite and scapolite, principally associated with breccia facies and mafic igneous units at the level of Upper Roan Subgroup decoupling (Selley et al., 2005). Sodic alteration locally extends into subsalt strata, primarily along layer-parallel shear zones nucleated along interfaces of contrasting rheology. Albitic alteration is associated with minor U-Mo-Cu mineralization but postdates authigenic K-feldspar and locally crosscuts stratiform Cu ores (Darnley, 1960; Darnley et al., 1961; J.L.W. Jolly, unpub. report, 1971; Lefèbvre and Tshiauka, 1986; Sweeney and Binda, 1989). Where the decoupling surface cuts down to the level of Lower Roan Subgroup synrift strata on the western flank of the Kafue anticline, these rocks are also intensely albitized. This configuration characterizes the synrift package-hosted Chibuluma and Chibuluma West deposits, where albite, intimately intergrown with Cu-Co sulfides, is the sole feldspathic phase (Darnley, 1960).

A somewhat similar configuration is apparent in the Domes region, where scapolite ± plagioclase are developed in shear zones along the basement-Lower Roan Subgroup contact (Cosi et al., 1992), and albite and scapolite occur in Upper Roan Subgroup breccia facies (Capistrant et al., 2015). Cosi et al. (1992) noted a ubiquitous association between shear-hosted U mineralization and scapolite, the latter having stabilized late relative to talc in basement rocks. Fluid inclusion data from U-associated quartz veins indicate they were deposited from moderate- to high-temperature, high-salinity NaCl-CaCl2 brines (Eglinger et al., 2014).

Kansanshi (Domes region) and Frontier (Zambian Copperbelt) are major suprasalt Cu deposits that are unequivocally associated with sodic metasomatism. In both deposits, ores are hosted by quartz-calcite ± dolomite veins with broad albitic halos (Torrealday et al., 2000; Broughton et al., 2002; Hitzman et al., 2012). The oldest Cu-bearing veins at the lower Nguba Group-hosted Kansanshi deposit crosscut the dominant tectonic fabric in the host rocks, and albite alteration overprinted metamorphic biotite. Albitic veins of the Frontier deposit are restricted to Mwashya Subgroup shales; the overlying Grand Conglomerate is unaltered and unmineralized (Hitzman et al., 2012). In the footwall to ore, Hitzman et al. (2012) show Mwashya Subgroup strata positioned directly on the Lower Roan synrift package, with the originally intervening Upper Roan Subgroup carbonates apparently removed along the decoupling surface. Unlike those at Kansanshi, however, the stockwork and albitic halos at Frontier are folded, with sulfides locally remobilized into the tectonic foliation. The contrasting timing of alteration relative to fabric development suggests the multistage or diachronous introduction of sodic fluids in the southern parts of the basin.

The suprasalt Kipushi deposit possesses an apparently unique alteration signature. Although developed on the margin of a diapiric breccia, which has the Mg signature typical of Congolese Copperbelt intrasalt packages (i.e., talc and chlorite), the fault- and vein-hosted orebodies are associated with a multistage assemblage of dolomite, quartz, Ba feldspar, Ba muscovite, Mg chlorite, phlogopite, and muscovite (Chabu and Boulegue, 1992; Heijlen et al., 2008).

Elemental analysis of fluid inclusions from Central African Copperbelt ores

Sodium-Cl-Br systematics from fluid inclusions have been shown in numerous studies to provide important constraints on the evolution of basinal brines (e.g., McCaffrey et al., 1987; Walter et al., 1990; Kesler et al., 1995; Yardley and Banks, 1995; Hammerli et al., 2013). In most cases, a bulk crush-leach method has been applied, and, while mixing of different fluid inclusion phases is inevitable, it is possible to discriminate between brines generated through progressive evaporation of seawater (i.e., residual brines) and those related to dissolution of marine evaporites. The approach relies on the fact that Br tends to remain conservative throughout the evaporation process. Thus, as evaporation proceeds, residual brines will have progressively lower Cl/Br values. By contrast, brines generated exclusively from the dissolution of halite will show a progression toward exceedingly high Cl/Br values from an initial seawater molar value of ~640.

Two previously published studies of this type relate to the ores of the Central African Copperbelt (Fig. 20a). Heijlen et al. (2008) present leachate analyses from authigenic carbonate and quartz of the Kipushi deposit, whereas Nowecki (2014) analyzed quartz and carbonate vein generations from a variety of deposits of the Zambian Copperbelt and Domes region. Both studies included conventional fluid inclusion analyses, which indicate ubiquitously high salinities >30 wt % NaCl equiv. Kipushi data appear distinct, with molar Cl/Br values consistently >640, indicative of salinity sourced predominantly from dissolution of halite. Zambian Copperbelt and Domes region molar Cl/Br values range from slightly higher than seawater to ~100, the latter corresponding to a residual brine that had evolved beyond the halite and carnallite saturation points (Nowecki, 2014). Veins associated with the classic Zambian Copperbelt stratiform Cu-Co ores possess the lowest molar Cl/Br values and conform most closely to the seawater evaporation trend. Domes region data show partial overlap with the Zambian Copperbelt data but with molar Cl/Br values closer to seawater and greater deviation from the seawater evaporation trend. This relationship led Nowecki (2014) to conclude that there was a temporal shift from highly evolved residual brine to probable synorogenic evaporite dissolution, the latter stage inclusions recording Na depletion via fluid-rock interaction and precipitation of albite and scapolite. This interpretation was supported by apparently lower molar Cl/Br values from vein generations considered to be pre- or synkinematic and higher molar Cl/Br values from postkinematic generations.

Fig. 20.

Na-Cl-Br variations in leachate data from various Central African Copperbelt deposit types and additional data from the Kupferschiefer. Cl/Br and Na/Br are molar ratios. (a) Previously published Central African Copperbelt data: Kipushi data sourced from Heijlen et al. (2008), the remainder from Nowecki (2014). (b) Newly presented data from the Congolese Copperbelt, Zambian Copperbelt, Domes region, and Kupferschiefer. (c) Expanded view of (b) highlighting positions of data relative to the seawater evaporation trend. Abbreviations: CCB = Congolese Copperbelt, NaCli = initial halite precipitation, NaClf = final halite precipitation, PGE = platinum group element, SET = seawater evaporation trend, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 20.

Na-Cl-Br variations in leachate data from various Central African Copperbelt deposit types and additional data from the Kupferschiefer. Cl/Br and Na/Br are molar ratios. (a) Previously published Central African Copperbelt data: Kipushi data sourced from Heijlen et al. (2008), the remainder from Nowecki (2014). (b) Newly presented data from the Congolese Copperbelt, Zambian Copperbelt, Domes region, and Kupferschiefer. (c) Expanded view of (b) highlighting positions of data relative to the seawater evaporation trend. Abbreviations: CCB = Congolese Copperbelt, NaCli = initial halite precipitation, NaClf = final halite precipitation, PGE = platinum group element, SET = seawater evaporation trend, Sgp = Subgroup, ZCB = Zambian Copperbelt.

We build upon these data sets with leachate analyses from a variety of Congolese Copperbelt, Zambian Copperbelt, and Domes region deposit types. These include (1) classic stratiform Cu-Co deposits of the Congolese Copperbelt (i.e., intrasalt, basin margin; Kolwezi, Kamoya, Kinsanfu, Etoile), (2) classic stratiform (arenite-hosted) Zambian Copperbelt Cu-Co ores (i.e., subsalt, basin margin; Chibuluma West, Nchanga), (3) basal Nguba Group-hosted vein-controlled Cu (Mo-Au) (i.e., suprasalt, basin margin; Kansanshi), and (4) Roan Group-hosted Cu U-Ni-Co (Au-PGE) (i.e., intrasalt, basin interior; Shinkolobwe). Supplementing this data set are leachates from the Polish Kupferschiefer, which also occupy a subsalt position in the Permian Zechstein basin (Hitzman et al., 2005). Results are given in Table 1 and plotted in Figure 20b and c. The majority of analyses are derived from disseminated sulfide and/or sulfide-gangue (e.g., dolomite, magnesite, quartz) mixtures, but included are dolomite matrix to breccia facies and veins that either transgress stratiform sulfides or form the principal ore host (e.g., Kansanshi).

Table 1.

Compositions of Leachates from Various Central African Copperbelt Deposit Types and Kupferschiefer Ore

SampleRegionDeposit typeDepositMineralNa (ng)NH4 (ng)K (ng)Mg (ng)Ca (ng)Sr (ng)Ba (ng)Cl (ng)Br (ng)SO4 (ng)Cat/AnCl/BrNa/Br
NS137 cpyZCBClassic arenite-hosted Cu-CoChibuluma WestCcp-Crl-Dol81.105.118.735.4717.46ndnd74.920.81204.000.83209349
NS137 1281 cpy/pyZCBClassic arenite-hosted Cu-CoChibuluma WestPy-Ccp139.775.1448.254.09185.910.622.01286.061.02457.590.98631475
NS137 1300.5ZCBClassic arenite-hosted Cu-CoChibuluma WestCcp-Crl108.841.8210.072.441.62ndnd111.001.1584.191.08217328
NC-car (NW)ZCBClassic arenite-hosted Cu-CoNchangaCrl369.8116.71176.371.7829.410.673.93875.794.73232.140.77417272
KN18 1163.3ZCBClassic argillite-hosted Cu-CoKonkola NorthDol272.431.9571.41106.31195.254.562.81591.993.4982.850.66383272
Mindola dupZCBClassic argillite-hosted Cu-CoMindolaDol-Anh121.731.16104.2783.95373.821.65nd209.561.64536.801.09289259
E1099BX14CCBClassic Cu-CoEtoileCct (in vein)348.012.3117.601.741.85nd1.98500.891.0964.171.031,0331,107
KYA219cpyCCBClassic Cu-CoKamoyaCcp209.216.83198.7440.89473.5119.8433.661,548.9812.94400.110.8027056
H3410 carCCBClassic Cu-CoKisanfuCrl217.365.27374.7633.30195.6813.0439.311,421.3811.220.000.6728667
KOV558/294.6CCBClassic Cu-CoKolweziCrl104.710.6742.1740.08198.614.7520.18673.555.6953.410.7926764
Luilu22 213CCBClassic Cu-CoKolweziCrl-Bn88.301.2053.8373.9570.855.4136.78416.982.3939.931.22394129
KTO559 421.5CCBClassic Cu-CoKolweziCrl-Bn48.790.1733.760.7838.171.770.65248.841.4413.100.68390118
KTO565 543.9CCBClassic Cu-CoKolweziCct136.170.2651.250.15104.816.597.38861.396.731.910.5228870
KTO565 553.9CCBClassic Cu-CoKolweziMixed sulfide98.541.05111.5539.30458.1611.2462.24748.024.27261.650.6239580
KTO565 548CCBClassic Cu-CoKolweziMixed sulfide412.971.9864.3432.19130.143.284.88996.512.9655.600.60758485
KTO565 543.9 dol/sulCCBClassic Cu-CoKolweziDol-Cct-Crl96.080.84125.8584.59420.834.0751.29767.465.7131.821.3330359
KADI19 165.9CCBClassic Cu-CoKolweziDol-Oz-sulfide845.407.27709.65122.041,682.1571.06155.356,918.3550.05126.150.6631259
KADI9 165.9CCBClassic Cu-CoKolweziDol-sulfide695.176.50622.7780.491,313.0056.71116.984,947.2339.71108.450.7328161
KADI19 165.9CCBClassic Cu-CoKolweziDol-sulfide1,081.339.54980.05171.591,986.3090.41170.577,638.8252.83101.910.8732671
KADI19 165.9 (NW)CCBClassic Cu-CoKolweziDol-Qz-sulfide590.745.17593.07158.851,155.5945.5222.424,332.1134.2662.040.8428560
DD030 221.56 magCCBClassic Cu-CoKinsanfuMgs3.370.5518.40139.49178.811.100.2320.340.1610.260.6228372
KTO565 553.1CCBClassic Cu-CoKolweziMgs162.072.37130.49185.38422.6811.87111.661,160.897.20128.520.5736478
KOV558 199CCBClassic Cu-CoKolweziDol (in breccia)57.250.482.6112.0657.0820.004.29138.290.3347.540.52936598
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiOz252.782.9022.96ndndnd15.39471.052.8717.810.39370307
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiOz1,436.109.85139.5663.79237.59nd18.402,342.8722.2095.361.23238225
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiCcp30.093.234.543.3528.71ndnd63.250.19149.040.67768564
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiCcp214.541.4617.1523.93557.90nd14.48445.902.66522.571.70378281
K343_214.8DomesVein-hosted Cu (Mo-Au)KansanshiPy207.951.3118.766.17109.49nd4.56431.391.90111.421.07512381
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiCal271.622.4539.1735.41378.57nd24.57581.152.9449.590.65445321
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiCal723.883.98108.83117.13197.53nd5.471,936.9412.7930.020.60341197
PZ-Ba calcitePolandStratiform CuKupferschieferCal734.833.8923.6558.34900.26nd6.412,254.6519.5991.531.26259130
PZ-vein cpyPolandStratiform CuKupferschieferCcp382.052.1213.4524.25481.31nd5.251,287.0813.19213.941.05220101
PZ-vein bornitePolandStratiform CuKupferschieferBn1,203.396.5244.1971.75888.97nd3.803,755.5636.10405.240.91235116
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweCrl463.644.18123.5338.79373.8519.5422.991,653.699.38172.410.86397172
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweCrl613.265.99150.0433.44722.6922.4116.912,209.818.51697.580.91585251
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol35.830.003.6229.5840.960.390.4388.680.1210.810.411,6091,003
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol50.480.153.4932.3633.200.480.50101.140.1912.890.551,221940
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol40.170.001.8559.2545.341.690.2981.68nd18.260.53  
SampleRegionDeposit typeDepositMineralNa (ng)NH4 (ng)K (ng)Mg (ng)Ca (ng)Sr (ng)Ba (ng)Cl (ng)Br (ng)SO4 (ng)Cat/AnCl/BrNa/Br
NS137 cpyZCBClassic arenite-hosted Cu-CoChibuluma WestCcp-Crl-Dol81.105.118.735.4717.46ndnd74.920.81204.000.83209349
NS137 1281 cpy/pyZCBClassic arenite-hosted Cu-CoChibuluma WestPy-Ccp139.775.1448.254.09185.910.622.01286.061.02457.590.98631475
NS137 1300.5ZCBClassic arenite-hosted Cu-CoChibuluma WestCcp-Crl108.841.8210.072.441.62ndnd111.001.1584.191.08217328
NC-car (NW)ZCBClassic arenite-hosted Cu-CoNchangaCrl369.8116.71176.371.7829.410.673.93875.794.73232.140.77417272
KN18 1163.3ZCBClassic argillite-hosted Cu-CoKonkola NorthDol272.431.9571.41106.31195.254.562.81591.993.4982.850.66383272
Mindola dupZCBClassic argillite-hosted Cu-CoMindolaDol-Anh121.731.16104.2783.95373.821.65nd209.561.64536.801.09289259
E1099BX14CCBClassic Cu-CoEtoileCct (in vein)348.012.3117.601.741.85nd1.98500.891.0964.171.031,0331,107
KYA219cpyCCBClassic Cu-CoKamoyaCcp209.216.83198.7440.89473.5119.8433.661,548.9812.94400.110.8027056
H3410 carCCBClassic Cu-CoKisanfuCrl217.365.27374.7633.30195.6813.0439.311,421.3811.220.000.6728667
KOV558/294.6CCBClassic Cu-CoKolweziCrl104.710.6742.1740.08198.614.7520.18673.555.6953.410.7926764
Luilu22 213CCBClassic Cu-CoKolweziCrl-Bn88.301.2053.8373.9570.855.4136.78416.982.3939.931.22394129
KTO559 421.5CCBClassic Cu-CoKolweziCrl-Bn48.790.1733.760.7838.171.770.65248.841.4413.100.68390118
KTO565 543.9CCBClassic Cu-CoKolweziCct136.170.2651.250.15104.816.597.38861.396.731.910.5228870
KTO565 553.9CCBClassic Cu-CoKolweziMixed sulfide98.541.05111.5539.30458.1611.2462.24748.024.27261.650.6239580
KTO565 548CCBClassic Cu-CoKolweziMixed sulfide412.971.9864.3432.19130.143.284.88996.512.9655.600.60758485
KTO565 543.9 dol/sulCCBClassic Cu-CoKolweziDol-Cct-Crl96.080.84125.8584.59420.834.0751.29767.465.7131.821.3330359
KADI19 165.9CCBClassic Cu-CoKolweziDol-Oz-sulfide845.407.27709.65122.041,682.1571.06155.356,918.3550.05126.150.6631259
KADI9 165.9CCBClassic Cu-CoKolweziDol-sulfide695.176.50622.7780.491,313.0056.71116.984,947.2339.71108.450.7328161
KADI19 165.9CCBClassic Cu-CoKolweziDol-sulfide1,081.339.54980.05171.591,986.3090.41170.577,638.8252.83101.910.8732671
KADI19 165.9 (NW)CCBClassic Cu-CoKolweziDol-Qz-sulfide590.745.17593.07158.851,155.5945.5222.424,332.1134.2662.040.8428560
DD030 221.56 magCCBClassic Cu-CoKinsanfuMgs3.370.5518.40139.49178.811.100.2320.340.1610.260.6228372
KTO565 553.1CCBClassic Cu-CoKolweziMgs162.072.37130.49185.38422.6811.87111.661,160.897.20128.520.5736478
KOV558 199CCBClassic Cu-CoKolweziDol (in breccia)57.250.482.6112.0657.0820.004.29138.290.3347.540.52936598
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiOz252.782.9022.96ndndnd15.39471.052.8717.810.39370307
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiOz1,436.109.85139.5663.79237.59nd18.402,342.8722.2095.361.23238225
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiCcp30.093.234.543.3528.71ndnd63.250.19149.040.67768564
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiCcp214.541.4617.1523.93557.90nd14.48445.902.66522.571.70378281
K343_214.8DomesVein-hosted Cu (Mo-Au)KansanshiPy207.951.3118.766.17109.49nd4.56431.391.90111.421.07512381
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiCal271.622.4539.1735.41378.57nd24.57581.152.9449.590.65445321
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiCal723.883.98108.83117.13197.53nd5.471,936.9412.7930.020.60341197
PZ-Ba calcitePolandStratiform CuKupferschieferCal734.833.8923.6558.34900.26nd6.412,254.6519.5991.531.26259130
PZ-vein cpyPolandStratiform CuKupferschieferCcp382.052.1213.4524.25481.31nd5.251,287.0813.19213.941.05220101
PZ-vein bornitePolandStratiform CuKupferschieferBn1,203.396.5244.1971.75888.97nd3.803,755.5636.10405.240.91235116
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweCrl463.644.18123.5338.79373.8519.5422.991,653.699.38172.410.86397172
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweCrl613.265.99150.0433.44722.6922.4116.912,209.818.51697.580.91585251
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol35.830.003.6229.5840.960.390.4388.680.1210.810.411,6091,003
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol50.480.153.4932.3633.200.480.50101.140.1912.890.551,221940
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol40.170.001.8559.2545.341.690.2981.68nd18.260.53  

Cl/Br and Na/Br are molar ratios; Cat/An is the ratio of total cation equivalents to anion equivalents

Abbreviations: Anh = anhydrite, Bn = bornite, Cal = calcite, CCB = Congolese Copperbelt, Ccp = chalcopyrite, Cct = chalcocite, Crl = carrollite, Dol = dolomite, Mgs = magnesite, nd = not determined, Py = pyrite, Qz = quartz, ZCB = Zambian Copperbelt

As shown in Figure 20, the compositional range of our data compares favorably with that of Nowecki (2014) but exhibits less deviation from the seawater evaporation trend. Leachates can be broadly divided into three groups based on their position relative to the seawater evaporation trend: (1) a group with molar Cl/Br values >640, indicative of salinity generated largely through halite dissolution, (2) a clustered data set with molar Cl/Br values from 220 to 400, positioned at or adjacent to the seawater evaporation trace, close to the transition from halite to bittern salt (i.e., Mg-dominant) precipitation, and (3) an intermediate group that broadly follows the trace of the seawater evaporation throughout the field of halite precipitation. Data indicative of halite dissolution include two of the four Shinkolobwe samples, three of 19 classic stratiform Congolese Copperbelt ores samples, including the single analyses of breccia facies and a transgressive vein, and an isolated sample from Kansanshi. Data for samples with disseminated ore textures from the classic stratiform Congolese Copperbelt and argillite-hosted Zambian Copperbelt ores plot in all but one case close to the halite-bittern salt transition (i.e., NaCl-MgSO4) and correspond to leachates from the Polish Kupferschiefer and the majority of Zambian Copperbelt stratiform Cu-Co ore data of Nowecki (2014). The relatively tight cluster of data is interpreted to indicate homogeneous residual brine, little affected by mixing with evaporite dissolution-related brines. The slight deviation of Congolese Copperbelt data from the seawater evaporation trend potentially records a history of Na depletion via precipitation of sodic phases, such as albite, prior to entrapment in ore sulfides. Alternatively, fluid-rock interactions involving the breakdown of Cl-bearing phases, such as replacement of clay or muscovite by Mg chlorite, may have increased Cl/Br values in the ore fluid. If this interpretation is correct, the unmodified composition of the residual brine would have been close to MgSO4 saturation, a scenario consistent with the intense Mg metasomatism that characterizes these ores.

Salinity sources for the third group of fluid inclusions with intermediate Cl/Br compositions are more equivocal. Including most Kansanshi data, arenite-hosted ores of the Zambian Copperbelt, and sulfide leachates from Shinkolobwe, the distribution is potentially indicative of mixing between residual and evaporite dissolution-related brines. This argument is perhaps most convincing for Kansanshi data, which collectively trend toward molar Cl/Br values >640 and conform closely to the seawater evaporation trend (e.g., Chi and Savard, 1997). Significant deviation of some arenite-hosted Zambian Copperbelt data (Chibuluma West) from the seawater evaporation trend again implies a component of fluid-rock interaction. Shinkolobwe data are particularly perplexing, as sulfide leachates (molar Cl/Br values 397–585) and carbonate leachates (molar Cl/Br values 1,221–1,609) from a single vein are significantly different. While a component of halite dissolution is probable in both leachate groups, distinct fluid compositions are likely to record multistage vein fill.

Despite some uncertainties with our data, which are expected utilizing a bulk-crush approach, the large variations in leachate compositions are best explained by input of, and in some cases mixing between, distinct fluids. Considering the entire data set (noting that data from Grand Conglomerate-hosted stratiform Cu-type ores are lacking), the following conclusions appear justified. First, a primary, highly evolved residual brine signature is characteristic of the giant stratiform intrasalt Congolese Copperbelt Cu-Co ores and the subsalt Polish Kupferschiefer. While the giant stratiform subsalt ores of the Zambian Copperbelt generally preserve a similar signature, there is evidence of mixing with evaporite dissolution-related brines in some of the arenite-hosted ores in particular. Second, in ores positioned toward the basin’s interior in the Congolese Copperbelt (i.e., Shinkolobwe) and in the Domes region, mixing of residual brine and evaporite dissolution-related brine appears ubiquitous. This signature is evident in stratiform Cu-Co ores primarily in transgressive veins and enclosing breccia facies. Third, a dominance of brine derived from halite dissolution occurs only at Kipushi in the Congolese Copperbelt-Domes region transition, the demonstrably youngest of the Central African Copperbelt ores. This fundamental change in salinity source matches the expected trend toward halite dissolution during later stages of a basin’s history. While Nowecki (2014) considered the transition to have occurred during orogenesis, as demonstrated in the following section, the behavior of salt during basin evolution is complex and varied; dissolution of salt potentially occurred diachronously and by different processes in different parts of the basin. In other words, brines that involved a component of evaporite dissolution do not necessarily record fluid migration that was ubiquitously younger than that associated with residual brines.

Implications for Basin Hydrology

The recognition of a residual brine source for most of the intra- and subsalt stratiform Central African Copperbelt Cu-Co deposits places important constraints on the hydrologic evolution of the Katangan basin. First, the production of residual brine in volumes sufficient to form the known ores requires deposition of thick evaporites. The only stage of basin development when this occurred was during middle Roan Group sedimentation at ~800 Ma (Fig. 21a). There is thus a ~250-m.y. gap between fluid production and onset of basin inversion. Second, the volume of residual brine within the basin is finite, as it can only be produced during periods of surficial evaporation. This is in contrast to brine formed through evaporite dissolution, which is limited only by the volume of salt and the flux of halite-undersaturated fluid. It is certain that the residual brine was diluted by interaction with pore fluids present in aquifers prior to evaporite deposition, which if halite undersaturated may have had little effect on Na-Cl-Br systematics; however, the high salinities associated with low Cl/Br-value fluid inclusions (>30 wt %; Nowecki, 2014) indicate that the volume of dilution was not significant.

Fig. 21.

Hydrologic model for the Katangan basin, showing permissible timings for the development of various alteration assemblages and types of mineralization, relative to stages of basin evolution. (a) First rift stage: deposition of Lower and Upper Roan Subgroups. Deposition of basin-wide salt sheets resulted in recharge of a basal arenite-dominated aquifer by an oxidizing Mg-rich residual evaporitic brine. Replacement of detrital phases by Mg chlorite in the depocenter maximum had the potential to initiate repartitioning of alkali elements (± light rare earth elements, Ni) toward basin-peripheral parts of the aquifer system. Possible early stage stratiform Cu-Co mineralization would have been restricted to the outer fringes of the salt sheets, where reducing elements, such as the Copperbelt Orebody Member in the Zambian Copperbelt, were in direct hydrological communication with the aquifer. (b) Mature stage of basin growth: deposition of the Nguba Group. Invigorated fluid flow in the basal aquifer was driven by increasing thermal contrast between the progressively subsiding depocenter maximum and the basin margins, contributing to the regional alteration zonation developed at subsalt levels. Primary hydrologic connectivity between the aquifer and reducing elements at the basin margins (including the Grand Conglomerate) demands that at least some component of the stratiform Cu (Co) ores would have formed at this time. Onset of halokinesis during Mwashya Subgroup and/or Grand Conglomerate deposition led to fragmentation of intrasalt strata and the potential for breaching of the basal salt seal at sites of welding. In basin marginal positions, where salt was originally thin, convective fluid cells had the potential to infiltrate intrasalt strata (i.e., multiple points of contact between intrasalt strata and the basal aquifer). This hydrologic framework is considered to have been fundamental in the formation of laterally extensive classic stratiform Congolese Copperbelt ores. The onset of salt dissolution potentially occurred during this period, with emergence of diapirs at the surface, or interaction of suprasalt, halite-undersaturated pores with domains of abnormally permeable salt; both of these conditions were most likely met in the depocenter maximum, where salt was originally thickest and most deeply buried. The hydrologic architecture depicted in center of the profile, involving connectivity of the basal aquifer to surface via a combination of welding and diapir collapse, may account for the mixed residual brine-halite dissolution-related brine Na-Cl-Br signature of Shinkolobwe solute data. (c) Orogenesis provided the strongest driving force for fluid flow in the basin’s history (Sillitoe et al., 2017a). At the initiation of basin inversion (noting that this may not have been synchronous at the basin scale), the fundamental hydrologic framework at the basin margins was likely little modified from early and/or intermediate stages of basin growth, allowing for formation and/or upgrading of existing, stratiform ores; however, effective fluid flow into the halokinetically dismembered intrasalt strata was, at this point, unlikely, with the possible exception of basin marginal zones where salt was originally very thin. As tectonism progressed and parts of the basin were deeply buried and metamorphosed salt dissolution pervaded the middle levels of the structural pile, leading to a fundamental reorganization of basin hydrology. Fluids derived from a mixture of residual- and salt-dissolution–related brine were directed along highly permeable former salt layers and diapirs to interact with suprasalt-reducing elements, forming fracture-controlled ore types. Partitioning of Na-dominant alteration phases largely within and above the level of dissolving salt is interpreted to reflect relatively high temperature fluid flow within the core of the orogen. Abbreviations: CCB = Congolese Copperbelt, COM = Copperbelt Orebody Member, Congl. = Conglomerate, Gp = Group, PGE = platinum group element, REE= rare earth element, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 21.

Hydrologic model for the Katangan basin, showing permissible timings for the development of various alteration assemblages and types of mineralization, relative to stages of basin evolution. (a) First rift stage: deposition of Lower and Upper Roan Subgroups. Deposition of basin-wide salt sheets resulted in recharge of a basal arenite-dominated aquifer by an oxidizing Mg-rich residual evaporitic brine. Replacement of detrital phases by Mg chlorite in the depocenter maximum had the potential to initiate repartitioning of alkali elements (± light rare earth elements, Ni) toward basin-peripheral parts of the aquifer system. Possible early stage stratiform Cu-Co mineralization would have been restricted to the outer fringes of the salt sheets, where reducing elements, such as the Copperbelt Orebody Member in the Zambian Copperbelt, were in direct hydrological communication with the aquifer. (b) Mature stage of basin growth: deposition of the Nguba Group. Invigorated fluid flow in the basal aquifer was driven by increasing thermal contrast between the progressively subsiding depocenter maximum and the basin margins, contributing to the regional alteration zonation developed at subsalt levels. Primary hydrologic connectivity between the aquifer and reducing elements at the basin margins (including the Grand Conglomerate) demands that at least some component of the stratiform Cu (Co) ores would have formed at this time. Onset of halokinesis during Mwashya Subgroup and/or Grand Conglomerate deposition led to fragmentation of intrasalt strata and the potential for breaching of the basal salt seal at sites of welding. In basin marginal positions, where salt was originally thin, convective fluid cells had the potential to infiltrate intrasalt strata (i.e., multiple points of contact between intrasalt strata and the basal aquifer). This hydrologic framework is considered to have been fundamental in the formation of laterally extensive classic stratiform Congolese Copperbelt ores. The onset of salt dissolution potentially occurred during this period, with emergence of diapirs at the surface, or interaction of suprasalt, halite-undersaturated pores with domains of abnormally permeable salt; both of these conditions were most likely met in the depocenter maximum, where salt was originally thickest and most deeply buried. The hydrologic architecture depicted in center of the profile, involving connectivity of the basal aquifer to surface via a combination of welding and diapir collapse, may account for the mixed residual brine-halite dissolution-related brine Na-Cl-Br signature of Shinkolobwe solute data. (c) Orogenesis provided the strongest driving force for fluid flow in the basin’s history (Sillitoe et al., 2017a). At the initiation of basin inversion (noting that this may not have been synchronous at the basin scale), the fundamental hydrologic framework at the basin margins was likely little modified from early and/or intermediate stages of basin growth, allowing for formation and/or upgrading of existing, stratiform ores; however, effective fluid flow into the halokinetically dismembered intrasalt strata was, at this point, unlikely, with the possible exception of basin marginal zones where salt was originally very thin. As tectonism progressed and parts of the basin were deeply buried and metamorphosed salt dissolution pervaded the middle levels of the structural pile, leading to a fundamental reorganization of basin hydrology. Fluids derived from a mixture of residual- and salt-dissolution–related brine were directed along highly permeable former salt layers and diapirs to interact with suprasalt-reducing elements, forming fracture-controlled ore types. Partitioning of Na-dominant alteration phases largely within and above the level of dissolving salt is interpreted to reflect relatively high temperature fluid flow within the core of the orogen. Abbreviations: CCB = Congolese Copperbelt, COM = Copperbelt Orebody Member, Congl. = Conglomerate, Gp = Group, PGE = platinum group element, REE= rare earth element, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Assuming the ore fluid responsible for the stratiform Central African Copperbelt Cu-Co deposits was dominated by a residual brine component, it is possible to calculate the minimum fluid and aquifer volumes involved. Kesler et al. (1995) undertook a similar analysis for Mississippi Valley-type ores of the United States, leading to estimated aquifer volumes in the order of 103 km3. Here, we consider the case of the classic Congolese Copperbelt ores. Taylor et al. (2013) compiled mineral resource figures from the Central African Copperbelt (166 Mt total premining Cu endowment) and indicated that known classic stratiform ores from the central and northern Congolese Copperbelt possess a premining endowment of 55.6 Mt Cu. This figure, although suitable for our purposes, does not include subore-grade Cu, nor the amount of metal either lost to erosion or preserved at depth. Moreover, it does not include resources of the Kamoa district, for which Ivanhoe Mines (corporate website, accessed March 2018) quote an indicated and inferred value of 36.6 Mt Cu. The latter is not included in our calculations, as we have no leachate data from Kamoa to constrain salinity sources, but it serves to illustrate that the 55.6 Mt Cu figure is a gross underestimate of the original amount of metal in the region, which must have been in excess of 100 Mt Cu.

The likely Cu content of the mineralizing fluid is not well constrained. Modern evaporitic brines have Cu concentrations up to ~20 ppm (e.g., White et al., 1963; Hartmann, 1985), whereas previously published estimates of ore fluid Cu content for Kupferschiefer and Katangan systems are 127 and 100 ppm, respectively (Cathles et al., 1993; Hitzman, 2000); a value of 100 ppm Cu is applied in this analysis. Accounting for the 55.6 Mt Cu resource figure and assuming a brine density of 1.15 g/cm3 (halite-saturated brine at ~150°C; e.g., Thurmond et al., 1984), a Cu content of 100 ppm, and an unlikely 100% efficiency of metal precipitation at all trap sites, the volume of ore fluid is 4.84 × 1011 m3. Assuming a reservoir porosity of 10% (i.e., sandstone at 3-km depth; Ramm, 1991), the reservoir volume is calculated to be 4,835 km3. This figure corresponds broadly to an estimated ~5,500 km3 preerosional volume of intensely Mg-metasomatized intrasalt arenites stratigraphically positioned below Mines Subgroup-hosted ore in the central and northern Congolese Copperbelt: x = longitudinal width of the central Congolese Copperbelt (~170 km), y = latitudinal width of the northern and central Congolese Copperbelt (~70 km)/orogenic stretch (0.65; Fig. 16), z = maximum preserved stratigraphic thickness of subore arenaceous strata (300 m).

As noted above, the calculated fluid and reservoir volumes are significant underestimates. Realistically, a much greater reservoir volume must have included stratigraphic intervals below the level of salt, namely the largely unexposed (in the Congolese Copperbelt) Lower Roan Subgroup synrift strata (Fig. 21). Storage of large volumes of brine below the level of the evaporite deposits is particularly attractive, because residual brine is dense and should have displaced presumably lower salinity connate fluids in preevaporitic stratigraphy.

The interpretation of a volumetrically significant reservoir of residual brine below the level of salt raises the likelihood of hydrologic connection between the Congolese Copperbelt and the Zambian Copperbelt. While the similar leachate compositions from classic stratiform Cu-Co deposits throughout the Central African Copperbelt support this conclusion, the apparent lateral transition from Mg-metasomatic stratabound assemblages developed in association with the classic Congolese Copperbelt Cu-Co ores to predominantly K-metasomatic assemblages in the Zambian Copperbelt (and indeed Kamoa to the north) requires further explanation. Although it is permissible that K metasomatism was in part, at least, residual brine derived, Na-Cl-Br systematics suggest a fluid composition that was likely Mg saturated as opposed to K saturated, a condition which is achieved at greater degrees of evaporation (Warren, 1999). Furthermore, Nowecki (2014) observes that K/Br values of Zambian Copperbelt (and also Domes region) leachates are typically higher than that achievable via seawater evaporation and suggests significant K enrichment through fluid-rock interaction. In this respect, the intense Mg metasomatism evident partly within, but principally below, the level of ore in Congolese Copperbelt, which occurred at the near-complete expense of primary alkali contents, provided a potential source of K. In the thickest part of the basin (e.g., Shinkolobwe), Mg metasomatism occurs throughout the entire Mines Subgroup interval, whereas to the north in the region of the classic Congolese Copperbelt Cu-Co ores, it is restricted to the lower part of the Mines Subgroup and older strata. The lateral transition to predominant K-metasomatic assemblages at the extreme basin peripheries (i.e., Kamoa and Zambian Copperbelt) thus imparts a broadly symmetrical zonation pattern, consistent with lateral expulsion from a major fluid reservoir positioned largely below the level of salt (Fig. 21). If so, it is permissible that the salt-peripheral Mwashya Subgroup aquifer positioned at the base of the Kamoa ore had been hydrologically connected to Mg-dominant subsalt aquifers underlying the classic Mines Subgroup-hosted ores.

It is tempting to include the intensely Mg-metasomatized interval below the level of salt-related decoupling in the Domes region within the same regionally extensive aquifer system. Indeed, the apparent preservation of primary alkali contents within the package suggests that it occupies an intermediate position along the inferred outflow path from the alkali-depleted Congolese Copperbelt and alkali-enriched Zambian Copperbelt; however, leachate data from ores contained within this compartment (i.e., Lumwana, Sentinel, and Enterprise; Fig. 20a) indicate that an evolved residual brine signature is rare, with most defining a trend toward Cl/Br values in excess of 640, suggestive of halite dissolution. It may be argued that fluid inclusions would not have preserved primary residual brine signatures through the high-grade metamorphic conditions experienced in the Domes region; however, we must consider the possibility that the stratabound Mg-metasomatic interval of the Domes region records a hydrological history different from that of subsalt strata in neighboring depocenters.

Hydrologic properties of salt

Halite is known for its extremely low permeability and its consequent ability to compartmentalize the hydrologic architecture of sedimentary basins (Warren, 1999). This behavior is evident in the immense stratabound alteration zones at sub- and intrasalt levels in the Katangan basin and in the relative paucity of stratabound Mg and/or K alteration at higher levels in the stratigraphy (except where primary salt sheets were absent; e.g., Kamoa). A dense residual brine appears to have been isolated within a stacked series of Lower Roan Subgroup aquifers (the basal synrift package being predominant) from the onset of evaporite deposition at ~800 Ma (Fig. 21a). Numerical fluid modeling studies of sedimentary basins suggest that fluid circulation within the aquifer inevitably commenced from these initial stages, driven by a variety of passive processes, including thermal buoyancy, asymmetric compaction, and density contrasts of brine and low-salinity connate waters (e.g., Sarkar et al., 1995; Swenson et al., 2004; Koziy et al., 2009). These processes all were likely to have operated until permeability was reduced to a point where fluid flow was inhibited. Only then would external (e.g., tectonic) forces resulting in compressed geothermal gradients and/or fluid overpressure be essential to continued fluid circulation in the basin. It seems unavoidable that some metal precipitation occurred at lower levels in the basin following accumulation of the salt (e.g., poorly constrained Re-Os ages of 816 ± 62 and 821 ± 51 Ma, 820 ± 62 and 762 ± 33 Ma from Zambian and Congolese Copperbelt stratiform Cu-Co ores, respectively; D. Broughton and F. Barra in Selley et al., 2005; Muchez et al., 2015).

During ongoing sediment accumulation and burial, the porosity and permeability of aquifer rocks are progressively reduced, limiting the capacity for fluid circulation in the basin. Conventional porosity/depth trends indicate that even in relatively coarse grained sandstones, porosity may be reduced to 5% at 4-km burial (e.g., Ramm and Bjørlykke, 1994); however, if conditions are such that pore fluid pressures approach lithostatic values, porosities >10% may theoretically be maintained to depths >6 km (Ramm, 1991), comparable to the maximum thickness of Nguba Group strata in the Congolese Copperbelt. In addition to limiting mechanical collapse of pore space, fluid overpressures also have the capacity to inhibit porosity-occluding mineral cementation, as argued by Stricker et al. (2016) to account for preservation of porosities up to 20% in reservoir sandstones buried to ~4.3 km in North Sea hydrocarbon fields. While we have no direct evidence of suprahydrostatic fluid pressures associated with burial in the Katangan basin, abnormal fluid pressures are characteristic of evaporite basins, particularly below laterally extensive salt seals (Warren, 1999). The near-basin-wide distribution of Roan Group salt revealed by our structural analysis permits the subsalt reservoir to have remained open to fluid flow until very advanced stages of growth, possibly until the onset of basin inversion.

Transient nature of the salt aquiclude: Abnormal fluid pressures were likely maintained within and below the level of salt until this seal was breached, either by physical or chemical processes. This is most likely to have occurred during advanced stages of basin evolution. It could have been driven by (1) extension-related sediment loading on the top of salt, (2) fluid overpressuring below the seal, (3) increasing temperature and pressure, and/or (4) dissolution due to interaction with large volumes of halite-undersaturated fluid from above (Fig. 21b, c).

Perhaps the most effective means of breaching the salt seal is via lateral withdrawal at the base of downward-impinging minibasins, i.e., welding (Fig. 21b). At the point of welding, the underlying reservoir is rapidly depressurized, expelling pore fluids to suprasalt levels. This mechanism is interpreted to be fundamental to fluid ingress to Mines Subgroup strata, which would otherwise have been hydrologically isolated from the main synrift reservoir. Evidence has been presented that potentially links the highest-grade parts of the two largest Mines Subgroup-hosted ore systems, Kolwezi and Tenke, with salt welds. Both possess anomalously high positive longitudinal strains and occupy crude synformal cores, the latter, in the case of Kolwezi, having been inverted during subsequent shortening to form a broad antiformal thrust stack. The synformal cores are considered to be inherited from the form of salt withdrawal minibasins, presently strongly inverted and largely or completely lost to erosion. In this interpretation, input of fluid from the subsalt reservoir occurred at the structurally lowest level of the pile; the roles of subsequent entrainment into diapiric heads or thrust emplacement to the top of the pile are simply to reposition the ores to economically viable crustal levels.

Significantly, the trigger for welding can occur in the absence of any imposed tectonic forces and is dependent on the interplay between rate and volume of sediment input and thickness of parent salt. For example, high sediment flux into the depocenter maximum may have led to very rapid amplification of diapirs and attenuation of salt below neighboring withdrawal basins, leading to preferential welding in this region. Conversely, differential loading of originally thin parent salt may have focused welds in basin marginal positions. Indeed, the onset of welding may have been diachronous throughout the basin, leading to the potential for ores with the same fundamental stratigraphic and compositional traits to have been formed at significantly different times. Given that the first evidence for salt withdrawal below Nguba Group-cored minibasins is during deposition of the Grand Conglomerate, this would appear to mark the earliest possible time for weld-related mineralization. This temporal constraint thus accords with the maximum age limit for ores hosted in basal Grand Conglomerate strata (e.g., Kamoa) positioned at the basin peripheries.

Salt modification via dissolution may occur either at the surface, through interaction of surface waters and emergent diapiric heads (Fig. 21b, c; Warren, 1999), or at depth, where increasing temperatures and/or pressure lead to changes in crystal structure that significantly increase the permeability of halite (Fig. 21c; Lewis and Holness, 1996; Kukla et al., 2011). In the former scenario, generation of dense, downward-percolating dissolution-related brines will not necessarily interact with and modify the composition of preexisting residual brines below the level of parent salt. In other words, spatially separate brine reservoirs of different origin may evolve simultaneously and independently, so long as the integrity of the basal salt seal is preserved.

Episodic mixing of the two brine types potentially accounts for the contrasting leachate compositions recorded in the Shinkolobwe sample. A combination of high-amplitude diapir emergence and salt welding at depth within the thickest part of the basin provides an explanation for the 650 Ma uraninite age for the deposit (Fig. 21b), which does not correspond with any defined external driving force (i.e., rifting or compression; e.g., Sillitoe et al., 2017a).

At temperatures above 300°C, or where pore pressures approach lithostatic values, salt has the potential to become permeable through subtle changes in the crystallographic form of halite or hydrofracturing, respectively (Lewis and Holness, 1996; Warren, 1999; Kukla et al., 2011). These important permeability transitions mean that salt layers can change from aquicludes to aquifers. The effect is magnified if proximal reservoirs of halite-undersaturated fluid contribute to dissolution. In the Katangan basin, elevated temperatures and/or pressures were likely to have existed (1) deep within the depocenter maximum at advanced stages of basin development, (2) immediately below the salt seal prior to breaching, (3) in the vicinity of intrusions (e.g., ~765 Ma mafic intrusions emplaced within evaporitic strata), and (4) during orogenesis. The elevated temperatures and pressures that likely existed deep within the depocenter maximum thus provide an alternative explanation for the heterogeneous leachate signature from Shinkolobwe; however, given the maximum burial depth of Mines Subgroup strata at the time of 650 Ma uraninite growth was ~6 km (i.e., maximum thickness of ~740–630 Ma Nguba Group strata), it is unlikely that temperatures exceeded 300°C during these earliest stages of U mineralization. Furthermore, while hydrofracturing of salt above a fluid reservoir with hydrostatic pore pressures was possible, a source of halite-undersaturated fluid would still have been required to cause salt dissolution.

If the mixed residual halite dissolution-related brine leachate signature from the subsalt compartment in the Domes region is considered a robust indicator of fluid source(s), it provides possible evidence of continued and/or rejuvenated stratabound Mg metasomatism during orogenic burial. The combination of thermal and/or pressure-controlled, grain-scale mechanical degradation of halite and release of a halite-undersaturated fluid component during metamorphic dehydration of detrital and authigenic phases had the potential to modify a residual brine through progressively increasing volumes of halite dissolution-related brine (Fig. 21c).

In both the Zambian Copperbelt and Domes region, the abrupt stratigraphically controlled, up-section transition from K and/or Mg metasomatism to intense Na metasomatism within and principally above the level of primary salt is perhaps the most compelling evidence for a fundamental salt-facilitated change in the permeability framework (Fig. 21c). During Na metasomatism, primary salt had clearly become an aquifer, forming the locus of albite and scapolite growth; however, at Kipushi, where the salt-dissolution leachate signature is strongest, the associated alteration assemblage consists of Mg ± K ± Ba-bearing minerals. Similarly, the breccia facies cements associated with the classic Congolese Copperbelt Cu-Co ores, while recording halite dissolution leachate signatures, are predominantly dolomite and Mg chlorite. Accordingly, the sodic alteration of the Zambian Copperbelt and Domes region is not a unique signature of halite dissolution.

The mechanism that appears to have preferentially stabilized sodic phases over Mg ± K in the Domes region and Zambian Copperbelt is not easily taken into account; however, given the higher metamorphic grades in these areas, increased temperature is a likely contributing factor. This is consistent with the synorogenic timing for the bulk of the Na metasomatism, as indicated by the ubiquitous fracture-controlled character of albite-associated suprasalt ores, Frontier and Kansanshi, along with postpeak metamorphic age constraints for the latter. We are unaware of geochemical modeling studies for the basin brines at temperatures achieved in the Domes region and Zambian Copperbelt (i.e., >400°C); however, modeling of volcanic-hosted massive sulfide systems indicates that albite will be precipitated in preference to K-bearing minerals (i.e., muscovite and K-feldspar) as temperatures increase beyond 250°C (Schardt et al., 2001).

The association of albitization with the subsalt arenite-hosted Chibuluma West ores and the evidence for a salt dissolution-related brine component may also be taken into account by elevated temperatures. This deposit is unusual in the Zambian Copperbelt in that it is directly overlain by breccia facies that entrains large intrusive mafic bodies attributed to ~765 Ma rift phase (Selley et al., 2005). While it cannot be discounted that the Na metasomatism relates to synorogenic dissolution of salt, it possibly records an earlier intrusion-related stage, localized atop an aquifer that was dominated by residual brine.

Although locally elevated temperatures appear to be the most plausible explanation for the apparent partitioning of Na metasomatism in the southern part of the basin, this does not accord with interpretation above that subsalt Mg metasomatism in the Domes region was also, in part at least, syn-orogenic and thus occurred at high temperatures. Clearly further work is required to resolve this paradox, which could include geochemical modeling or leachate analysis of Mg and Na phases (as opposed to the existing data set largely derived from quartz ± carbonate veins).

Conclusions

Giant stratiform Cu ores are positioned in lateral outflow zones of peripherally thinning series of basal aquifers. The aquifer system was capped and compartmentalized by basin-wide primary salt sheets. In this respect, the Katangan basin is similar to the Polish Kupferschiefer, the world’s other great sediment-hosted stratiform Cu province. A residual brine signature, generated from the deposition of capping evaporites, is common to stratiform ores of both provinces and considered representative of ore fluids. In the Katangan basin, the pathway of the residual brine is recorded by a macroscopic zonation from Mg metasomatism, coupled with alkali element depletion, in the depocenter maximum and its northward-tapering margin in the Congolese Copperbelt to progressively increasing concentrations of potassium where the stratigraphy is most condensed.

Estimates of reservoir budgets for the central and northern Congolese Copperbelt permit part of the classic stratiform Cu-Co ores to have been derived from a subore aquifer, hydrologically isolated within a broader envelope of salt. Mineralization may have initiated from deposition of supra-Mines Subgroup salt (i.e., downward flux of residual brine to the aquifer), as evidenced by a ~800 Ma phase of sulfide precipitation at Kolwezi (Muchez et al., 2015). In the central Congolese Copperbelt, where salt was thicker and significantly affected by halokinesis, an upper age limit for this phase of mineralization is likely constrained to the Nguba Group depositional period, during which the Mines Subgroup and its basal aquifer were dismembered.

To account for the considerably greater metal budget of the northern Congolese Copperbelt, including the Kamoa domain, connectivity with the subsalt synrift aquifer was required. Predominance of the residual brine signature suggests the necessary change in hydrologic configuration was achieved via lateral withdrawal of salt as opposed to dissolution. Fluids expelled upward from the deep-seated aquifer interacted with relatively intact segments of Mines Subgroup strata trapped below downward-impinging Nguba Group-cored minibasins. Highest-grade mineralization occurred immediately above the point of welding, diminishing laterally to the points at which dismembered écailles became entrained in salt and were not in hydrological communication with the aquifer. Continued impingement of the minibasins led to lateral shortening of mineralized Mines Subgroup strata, the earliest stages of a complex, heterogeneous strain history.

Fluid focusing may have been enhanced where welding was coincident with perturbations in the synrift aquifer, such as abrupt thickness changes across growth faults. The spatial correspondence of high-grade mineralization at Tenke with the apparent northern pinch-out of Mines Subgroup strata and overlying minibasin cores supports the interpretation of a long-lived subbasin margin. Similarly, the correspondence of ore with abrupt Nguba Group thickness gradients about the fringe of the depocenter maximum (i.e., central Congolese Copperbelt) and projection of these trends into basin marginal positions, where they also coincide with mineralization (i.e., Kamoa), suggests a fundamental control of macroscopic subbasin architecture on ore location.

The earliest timing for weld-related fluid input to the Mines Subgroup is constrained by the acceleration of halokinesis during Grand Conglomerate deposition. Welding likely occurred diachronously across the basin, allowing for multiple mineralization ages, but most probably initiated where salt was originally thinnest. In areas where the parent salt (and the possible laterally emplaced salt allochthon above Kamoa) remained intact, it is possible that elevated pore pressures and fluid circulation were maintained until the latest stages of basin development, potentially through orogenesis. As long as the aquifer remained isolated from reducing elements in the stratigraphy, the ore fluid would have had the capacity to leach the entire reservoir of its detrital metal budget. Thus, it would appear that prolonged hydrological isolation was favorable to ore formation.

Although the onset of halokinetic disruption of the Mines Subgroup defines the youngest age constraint for classic stratiform Congolese Copperbelt ore formation, we cannot discount the possibility that in areas of thin salt, where the disruptive effects may have been less, stratiform mineralization persisted until the main stage of tectonic inversion. According to our structural model, the main phase of shortening in the northern Congolese Copperbelt postdated peak metamorphism and a phase of extensional collapse in the Domes region. Thus, a component of synorogenic ore formation appears permissible. Indeed, while we have reservations over whether leachate signatures from the subsalt Mg-metasomatized interval of the Domes region are robust, the data are consistent with thermally induced salt dissolution, suggesting the salt aquiclude was degrading but essentially remained intact during at least the earliest stages of orogenesis.

In the Zambian Copperbelt, the positioning of classic stratiform ores below the level of salt limited the necessity of the more complex hydrological evolution of the Congolese Copperbelt. The basal synrift aquifer, infiltrated at ~800 Ma by residual brine, was at all times in hydrological communication with at least the in situ reducing elements of the Copperbelt Orebody Member. Migration of preore-stage hydrocarbon to the sites of arenite-hosted ores requires a post-early diagenetic timing for this style of mineralization; however, as noted by Selley et al. (2005), the absolute age of hydrocarbon trapping is poorly constrained to the period of ~765 Ma rifting and associated heat input through to the onset of orogenesis. The most unequivocal age constraint remains the macroscopic geometry of ore, which suggests a prefolding/foliation timing for fluid input. Again, however, it is possible that foliation development in the Zambian Copperbelt postdated peak metamorphism in the Domes region, such that the ~500 Ma Re-Os molybdenite ages for mineralization (Sillitoe et al., 2017a) make geologic sense. Until robust ages for foliation development are determined, ambiguity surrounding mineralization timing will persist.

The onset of salt dissolution was likely to have commenced during extensional stages of basin development, particularly within the depocenter maximum where evaporites were sufficiently voluminous for high-amplitude diapirs to have become emergent. A halite-dissolution signature occurs in association with Shinkolobwe U-Ni-Co (Au-PGE) mineralization and distinguishes it from the classic Congolese Copperbelt Cu-Co ores. The contrasting fluid chemistries thus likely record different hydrological histories of the depocenter maximum and northern basin fringe and may have contributed to the macroscopic partitioning of ore types in these parts of the basin.

Synorogenic, albite-associated Cu ores of the Domes region and Zambian Copperbelt record a fundamental change in basin hydrology, with fluid flow having become focused within the evaporite stratigraphy and permeating the lowermost levels of the Nguba Group. The conditions that led to the apparent change from Mg ± K metasomatism to Na metasomatism remain equivocal, but a thermal control is favored. Although the strong partitioning of sodic alteration minerals within and at the peripheries of the former salt provided compelling evidence for salt dissolution, leachate data from Kansanshi indicate that the residual brine, possibly sourced from the depocenter maximum to the north, was still contributing to the ore fluid. The salt budget of the basin was not exhausted at the end of the orogenic period, as evidenced by the strong halite dissolution signature of ~450 Ma Kipushi ore. Considering the close spatial association of the latter with remnants of a high-amplitude diapir, we suspect that salt was partly preserved in the depocenter maximum where it either emerged at the surface or interacted with an influx of halite-undersaturated fluids, possibly during a period of orogenic collapse.

To summarize, the Katangan basin records a complex, likely progressive history of hydrological modification, largely attributable to the deposition and physiochemical properties of salt. The timing of certain modification stages that led to mineralization remains equivocal for several of the deposit types; however, this is unlikely to affect the chances of exploration success. Whether the timing of mineralization is considered diagenetic, synorogenic, or postorogenic, the location of most ore types, the giant stratiform varieties in particular, is fundamentally controlled by the form and distribution of macro-and deposit-scale redox and hydrological elements that were in place from the early stages of basin growth. Exploration models should focus on constraining this architecture and consider how the evolution of permeability networks through time could have led to the formation of known or new deposit types.

Acknowledgments

Much of the work presented here was undertaken as part of the AMIRA-ARC P544 and P872 projects. Not included in the author list are research staff and postgraduate students from the University of Tasmania (Peter McGoldrick, Ross Large, Mawson Croaker, Wallace Mackay, and Nicky Pollington) and the Colorado School of Mines (Nick Harris) who made important contributions to aspects of the projects not covered in this paper. We greatly appreciate the support and intellectual input of the industry-based sponsor group, without which this work would have been impossible. Thorough reviews of an earlier manuscript by Cliff Taylor, Jeffrey Mauk, an anonymous reviewer, and Janet Slate have significantly strengthened many aspects of this work. We are greatly indebted to them for their expert contributions. Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

REFERENCES

Annels
,
A.E.
,
1979
,
Mufulira greywackes and their associated sulphides
:
Institution of Mining and Metallurgy Transactions
 , v.
88
, p.
B15
23
.
Annels
,
A.E.
,
1989
,
Ore genesis in the Zambian Copperbelt, with particular reference to the northern sector of the Chambishi basin
:
Geological Association of Canada, Special Paper 36
 , p.
427
452
.
Armstrong
,
R.A.
,
Robb
,
L.J.
,
Master
,
S.
,
Kruger
,
F.J.
, and
Mumba
,
P.A.C.C.
,
1999
,
New U-Pb age constraints on the Katangan sequence, Central African Copperbelt [abs.]
:
Journal of African Earth Sciences, Special Issue
 , v.
28
, p.
6
7
.
Ball
,
S.M.
,
2015
,
Hydrothermal nickel sulfide hosted in Neoproterozoic carbonate and evaporitic rocks of the Menda Central prospect, Democratic Republic of Congo
: Unpublished M.Sc. thesis,
Golden, Colorado
,
Colorado School of Mines
,
124
p.
Barra
,
F.
,
Broughton
,
D.
,
Ruiz
,
J.
, and
Hitzman
,
M.W.
,
2004
,
Multi-stage mineralization in the Zambian Copperbelt based on Re-Os isotope constraints[abs.]
:
Geological Society of America Abstracts with Programs
 , v.
36
, p.
516
.
Barron
,
J.W.
,
2003
,
The stratigraphy, metamorphism, and tectonic history of the Solwezi area, Northwest Province, Zambia
:
Integrating geological field observations and airborne geophysics in the interpretation of regional geology
 : Unpublished Ph.D. thesis,
Golden, Colorado
,
Colorado School of Mines
,
233
p.
Bartholomé
,
P.
,
Evrard
,
P.
,
Katekesha
,
F.
,
Lopez-Ruiz
,
J.
, and
Ngongo
,
M.
,
1973
,
Diagenetic ore-forming processes at Kamoto, Katanga, Republic of the Congo
, in
Amstutz
,
G.C.
, and
Bernard
,
A.J.
, eds.,
Ore in sediments
 :
New York
,
Springer-Verlag
, p.
21
41
.
Batumike
,
M.J.
,
Cailteux
,
J.L.H.
, and
Kampunzu
,
A.B.
,
2007
,
Lithostratigraphy, basin development, base metal deposits, and regional correlations of the Neoproterozoic Nguba and Kundelungu rock successions
,
Central African Copperbelt: Gondwana Research
 , v.
11
, p.
432
447
.
Bernau
,
R.
,
Roberts
,
S.
,
Richards
,
M.
,
Nisbet
,
B.
,
Boyce
,
A.
, and
Nowecki
,
J.
,
2013
,
The geology and geochemistry of the Lumwana Cu (± Co ± U) deposits, NW Zambia
:
Mineralium Deposita
 , v.
48
, p.
137
153
.
Brems
,
D.
,
Muchez
,
Ph.
,
Sikazwe
,
O.
, and
Mukumba
,
W.
,
2009
,
Metallogenesis of the Nkana copper-cobalt South orebody, Zambia
:
Journal of African Earth Sciences
 , v.
55
, p.
185
196
.
Broughton
,
D.W.
, and
Rogers
,
T.
,
2010
,
Discovery of the Kamoa copper deposit, Central Africa Copperbelt, D.R.C.
:
Society of Economic Geologists Special Publication 15
 , v.
1
, p.
287
298
.
Broughton
,
D.W.
,
Hitzman
,
M.W.
, and
Stephens
,
A.J.
,
2002
,
Exploration history and geology of the Kansanshi Cu-(Au) deposit, Zambia
:
Society of Economic Geologists Special Publication
 
9
, p.
141
153
.
Bull
,
S.
,
Selley
,
D.
,
Broughton
,
D.
,
Hitzman
,
M.
,
Cailteux
,
J.
,
Large
,
R.
, and
McGoldrick
,
P.
,
2011
,
Sequence and carbon isotopic stratigraphy of the Neoproterozoic Roan Group strata of the Zambian copperbelt
:
Precambrian Research
 , v.
190
, p.
70
89
.
Byrne
,
G.M.
,
2017
,
Geometries and genesis of breccias within the halokinetic Kaulukundi diapir at the Mashitu mine, Katanga, Democratic Republic of Congo
: Unpublished MSc. thesis,
Golden, Colorado
,
Colorado School of Mines
,
146
p.
Cahen
,
L.
,
François
,
A.
, and
Ledent
,
D.
,
1971
,
Sur l’age des uraninites de Kambove Ouest et de Kamoto Principal et révision des connaissances relatives aux minéralisations uranifères du Katanga et du Copperbelt de Zambia
:
Annales de la Société Géologique de Belgique
 , v.
94
, p.
185
198
.
Cailteux
,
J.
,
1977
,
Particularites stratigraphiques et petrographiques du faiceau inferieur du Groupe des Mines au centre de l’arc cuperifere Shabien
:
Annales de la Société Géologique de Belgique
 , v.
100
, p.
55
71
.
Cailteux
,
J.
,
1986
,
Diagenetic sulphide mineralization within the stratiform coppercobalt deposit of West Kambove (Shaba-Zaire). Sequence of mineralization in sediment-hosted copper deposits
(part
2
), in
Friedrich
,
G.H.
, et al
, eds.,
Geology and metallogeny of copper deposits
 :
Berlin, Heidelberg
,
Springer-Verlag
, p.
398
411
.
Cailteux
,
J.
,
1991
,
La tectonique intra-Katanguienne dans la region nord-ouest de l’arc Lufilien (Shaba, Rép. Du Zaire)
:
Annales de la Société Géologique de Belgique
 , v.
113
, p.
199
215
.
Cailteux
,
J.
, and
Lefèbvre
,
J.J.
,
1975
,
Stratigraphie et minéralisations du gisement cuprifère de Kipapila, Shaba, Zaire
:
Annales de la Société Géologique de Belgique
 , v.
98
, p.
317
330
.
Cailteux
,
J.
,
Binda
,
P.
,
Katekesha
,
W.M.
,
Kampunzu
,
A.B.
,
Intiomale
,
M.M.
,
Kapenda
,
D.
,
Kaunda
,
C.
,
Ngongo
,
K.
,
Tshiauka
,
T.
, and
Wendorff
,
M.
,
1994
,
Lithostratigraphic correlation of the Neoproterozoic Roan Supergroup from Shaba (Zaïre) and Zambia, in the central African copper-cobalt metallogenic province
:
Journal of African Earth Sciences
 , v.
19
, p.
265
278
.
Cailteux
,
J.L.H.
,
Kampunzu
,
A.B.
,
Lerouge
,
C.
,
Kaputo
,
A.K.
, and
Milesi
,
J.P.
,
2005
,
Genesis of sediment-hosted stratiform copper-cobalt deposits, Central African Copperbelt
:
Journal of African Earth Sciences
 , v.
42
, p.
134
158
.
Capistrant
,
P.L.
,
Hitzman
,
M.W.
,
Wood
,
D.
,
Kelley
,
N.M.
,
Williams
,
G.
,
Zimba
,
M.
,
Kuiper
,
Y.
,
Jack
,
D.
, and
Stein
,
H.
,
2015
,
Geology of the Enterprise hydrothermal nickel deposit, North-Western Province, Zambia
:
Economic Geology
 , v.
110
, p.
9
38
.
Cathles
L.M.
,
Oszczepalski
,
S.
, and
Jowett
,
E.C.
,
1993
,
Mass balance evaluation of the late diagenetic hypothesis for Kupferschiefer Cu mineralization in the Lubin basin of southwestern Poland
:
Economic Geology
 , v.
88
, p.
948
956
.
Chabu
,
M.
, and
Boulègue
,
J.
,
1992
,
Barian feldspar and muscovite from the Kipushi Zn-Pb-Cu deposit, Shaba, Zaïre
:
Canadian Mineralogist
 , v.
30
, p.
1143
1152
.
Chi
,
G.
, and
Savard
,
M.M.
,
1997
,
Sources of basinal and Mississippi Valleytype mineralizing brines: Mixing of evaporated seawater and halite dissolution brine
:
Chemical Geology
 , v.
143
, p.
121
125
.
Cosi
,
M.
,
De Bonis
,
A.
,
Gosso
,
G.
,
Hunziker
,
J.
,
Martinotti
,
G.
,
Moratto
,
S.
,
Robert
,
J.P.
, and
Ruhlman
,
F.
,
1992
,
Late Proterozoic thrust tectonics, high-pressure metamorphism and uranium mineralization in the Domes area, Lufilian arc, northwestern Zambia
:
Precambrian Research
 , v.
58
, p.
215
240
.
Darnley
,
A.G.
,
1960
,
Petrology of some Rhodesian Copperbelt orebodies and associated rocks
:
Institute of Mining and Metallurgy Transactions
 , v.
69
, p.
137
173
.
Darnley
,
A.G.
,
Horne
,
J.E.T.
,
Smith
,
G.H.
,
Chandler
,
T.R.D.
,
Dance
,
D.F.
, and
Preece
,
E.R.
,
1961
,
Ages of some uranium and thorium minerals from east and central Africa
:
Mineralogical Magazine
 , v.
32
, p.
716
724
.
Decrée
,
S.
,
Deloule
,
É.
,
De Putter
,
T.
,
Dewaele
,
S.
,
Mees
,
F.
,
Yans
,
J.
,
Marignac
,
C.
,
2011
,
SIMS U-Pb dating of uranium mineralization in the Katanga Copperbelt
:
Constraints for the geodynamic context: Ore Geology Reviews
 , v.
40
, p.
81
89
.
de Magnée
,
I.
, and
François
,
A.P.
,
1988
,
The origin of the Kipushi (Cu, Zn, Pb) deposit in direct relation with a Proterozoic salt diapir; Copperbelt of Central Africa, Shaba, Republic of Zaire
:
Society for Geology Applied to Mineral Deposits, Special Publication
 
5
, p.
74
93
.
Demesmaeker
,
G.
,
François
,
A.
, and
Oosterbosch
,
R.
,
1963
,
La tectonique des gisements cuprifeères stratiformes du Katanga
, in
Lombard
,
J.
, and
Nicolini
,
P.
, eds.,
Gisements stratiformes de cuivre en Afrique, 2ème partie: Tectonique: Association des Services Géologiques Africains (Paris), Lusaka
 
1962
, p.
47
115
.
Derriks
,
J.J.
, and
Vaes
,
J.F.
,
1956
,
The Shinkolobwe uranium deposits: Ccurrent status of our geological and metallogenic knowledge
:
International Conference on the Peaceful Uses of Atomic Energy
,
United Nations
,
Geneva
, August 8–20, 1955, Proceedings, v.
6
, p.
94
128
.
Dewaele
,
S.
,
Muchez
,
Ph.
,
Vets
,
J.
,
Fernandez-Alonzo
,
M.
,
Tack
,
L.
,
2006
,
Multiphase origin of the Cu-Co ore deposits in the western part of the Lufilian fold-and-thrust belt, Katanga (Democratic Republic of Congo)
:
Journal of African Earth Sciences
 , v.
46
, p.
455
469
.
Duffett
,
M.
,
Sebagenzi
,
S.
, and
Selley
,
D.
,
2010
,
Gravity and magnetics of the Central African Copperbelt
:
AMIRA P872—Sediment-hosted copper deposits of Congolese, Zambian and central Australian basin systems, Final Report
 , p.
49
78
.
Eglinger
,
A.
,
Ferraina
,
C.
,
Tarantola
,
A.
,
André-Mayer
,
A.S.
,
Vanderhaeghe
,
O.
,
Boiron
,
M.C.
,
Dubessy
,
J.
,
Richard
,
A.
, and
Brouand
,
M.
,
2014
,
Hypersaline fluids generated by high-grade metamorphism of evaporites
:
Fluid inclusion study of uranium occurrences in the western Zambian Copperbelt: Contributions to Mineralogy and Petrology
 , v.
167
, p.
967
994
.
Eglinger
,
A.
,
Vanderhaeghe
,
O.
,
André-Mayer
,
A.S.
,
Goncalves
,
P.
,
Zeh
,
A.
,
Durand
,
C.
, and
Deloule
,
E.
,
2016
,
Tectono-metamorphic evolution of the internal zone of the Pan-African Lufilian orogenic belt (Zambia)
:
Implications for crustal reworking and synorogenic uranium mineralizations: Lithos
 , v.
240
, p.
167
188
.
El Desouky
,
H.A.
,
Muchez
,
P.
, and
Cailteux
,
J.
,
2009
,
Two Cu-Co phases and contrasting fluid systems in the Katanga Copperbelt
,
Democratic Republic of Congo: Ore Geology Reviews
 , v.
36
, p.
315
332
.
Fleischer
,
V.D.
,
Garlick
,
W.G.
, and
Haldane
,
R.
,
1976
,
Geology of the Zambian Copperbelt
, in
Wolf
,
K.H.
, ed.,
Handbook of stratabound and stratiform ore deposits
 : v.
6
, p.
223
352
.
François
,
A.P.
,
1974
,
Stratigraphie, teconique et minéralisations dans l’Arc cuprifére du Shaba (Rép. du Zaïre)
, in
Bartholome
,
P.
ed.,
Gisements striformes et provinces cupriféres: Centernaire de la Société Géologique de Belgique, Liège
 , p.
79
101
.
François
,
A.P.
,
1986
,
La parties centrale de l’Arc cuprifere du Katanga
:
étude géologique: Tervuren African Geoscience Collection
 , v.
109
,
61
p.
François
,
A.P.
,
1994
,
La structure tectonique du Katanguien dans la region de Kolwezi (Shaba, Rep. du Zaire)
:
Annales de la Société géologique de Belgique
 , v.
116
, p.
87
104
.
François
,
A.P.
, and
Cailteux
,
J.
,
1981
,
La Couverture Katangienne entre les Socles de Zilo et de la Kambompo, République du Zaire—région de Kolwezi
:
Tervuren, Belgique, Musée Royal de l’Afrique Centrale, Annales Serie In 8o, Sciences Géologiques, no
 .
87
,
50
p.
Fuchs
,
L.
,
Schmeling
,
H.
, and
Koyi
,
H.
,
2011
,
Numerical models of salt diapir formation by down-building: The role of sedimentation rate, viscosity contrast, initial amplitude, and wavelength
:
Geophysical Journal International
 , v.
186
, p.
390
400
.
Halley
,
S.W.
,
Wood
,
D.
,
Stoltze
,
A.
,
Godfroid
,
J.
,
Goswell
,
H.
, and
Jack
,
D.
,
2016
,
Using multielement geochemistry to map multiple components of a mineral system
:
Case study from a sediment-hosted Cu-Ni camp, NW Province, Zambia: Society of Economic Geologists Newsletter
 , no.
104
, p.
1
,
15
21
.
Halpin
,
J.
, and
Selley
,
D.
,
2010
,
Insights to geometry and structural evolution of the Central African Copperbelt from zircon provenance analysis
:
AMIRA P872—sediment-hosted copper deposits of Congolese, Zambian, and central Australian basin systems, Final Report
 , p.
33
47
.
Hammerli
,
J.
,
Rusk
,
B.
,
Spandler
,
C.
,
Emsbo
,
P.
, and
Oliver
,
N.H.S.
,
2013
,
In situ quantification of Br and Cl in minerals and fluid inclusions by LA-ICP-MS
:
A powerful tool to identify fluid sources: Chemical Geology
 , v.
337–338
, p.
75
87
.
Hanson
,
R.E.
,
Wardlaw
,
M.S.
,
Wilson
,
T.J.
, and
Mwale
,
G.
,
1993
,
U-Pb zircon ages from the Hook granite massif and Mwembeshi dislocation: Constraints on Pan-African deformation, plutonism, and transcurrent shearing in central Zambia
:
Precambrian Research
 , v.
63
, p.
189
209
.
Hartmann
,
M.
,
1985
,
Atlanis-II deep geothermal brine system
.
Chemical processes between hydothermal brines and Red Sea deep water: Marine Geology
 , v.
64
, p.
157
177
.
Heijlen
,
W.
,
Banks
,
D.A.
,
Muchez
,
P.
,
Stensgard
,
B.M.
, and
Yardley
,
B.W.D.
,
2008
,
The nature of mineralizing fluids of the Kipushi Zn-Cu deposit, Katanga, Democratic Republic of Congo
:
Quantitative fluid inclusion analysis using laser ablation ICP-MS and bulk crush-leach methods: Economic Geology
 , v.
103
, p.
1459
1482
.
Hendrickson
,
M.D.
,
Hitzman
,
M.W.
,
Wood
,
D.
,
Humphrey
,
J.D.
, and
Wendlandt
,
R.F.
,
2015
,
Geology of the Fishtie deposit, Central Province, Zambia
:
Iron oxide and copper mineralisation in Nguba Group metasedimentary rocks, Mineralium Deposita
 , v.
50
, p.
717
737
.
Hitzman
,
M.W.
,
2000
,
Source basins for sediment-hosted stratiform Cu deposits: Implications for the structure of the Zambian Copperbelt
:
Journal of African Earth Sciences
 , v.
30
, p.
855
863
.
Hitzman
,
M.W.
, and
Broughton
,
D.W.
,
2017
,
Discussion: “Age of the Zambian Copperbelt”
:
Mineralium Deposita
 , v.
52
, p.
1273
1275
.
Hitzman
,
M.W.
,
Kirkham
,
R.
,
Broughton
,
D.
,
Thorson
,
J.
, and
Selley
,
D.
,
2005
,
The sediment-hosted stratiform copper ore system
:
Economic Geology 100th Anniversary Volume
 , p.
609
642
.
Hitzman
,
M.W.
,
Broughton
,
D.
,
Selley
,
D.
,
Woodhead
,
J.
,
Wood
,
D.
, and
Bull
,
S.
,
2012
,
The Central African Copperbelt: Diverse stratigraphic, structural, and temporal settings in the world’s largest sedimentary copper district
:
Society of Economic Geologists Special Publication
 
16
, p.
487
514
.
Intiomale
,
M.M.
, and
Oosterbosch
,
R.
,
1974
,
Géologie et géochimie du gisement de Kipushi–Zaire
, in
Bartholome
,
P.
, ed.,
Gisements stratiformes et provinces cupifères: Centenaire de la Société Géologique Belgique de Liège
 , p.
123
164
.
Jackson
,
M.P.A.
,
Warin
,
O.N.
,
Woad
,
G.M.
, and
Hudec
,
M.R.
,
2003
,
Neoproterozoic allochthonous salt tectonics during the Lufilian orogeny in the Katangan Copperbelt, central Africa
:
Geological Society of America Bulletin
 , v.
115
, p.
314
330
.
John
,
T.
,
Schenk
,
V.
,
Mezger
,
K.
, and
Tembo
,
F.
,
2004
,
Timing and PT evolution of whiteschist metamorphism in the Lufilian arc-Zambezi belt orogen (Zambia)
:
Implications for the assembly of Gondwana: Journal of Geology
 , v.
122
, p.
71
90
.
Kampunzu
,
A.B.
, and
Cailteux
,
J.
,
1999
,
Tectonic evolution of the Lufilian arc (Central Africa Copper Belt) during Neoproterozoic Pan African orogenesis
:
Gondwana Research
 , v.
2
, p.
401
421
.
Kampunzu
,
A.B.
,
Tembo
,
F.
,
Matheis
,
G.
,
Kapenda
,
D.
, and
Huntsman-Mapila
,
P.
,
2000
,
Geochemistry and tectonic setting of mafic igneous units in the Neoproterozoic Katangan basin, central Africa: Implications for Rodinia break-up
:
Gondwana Research
 , v.
3
, p.
125
153
.
Kennedy
,
K.
,
Eyles
,
N.
, and
Broughton
,
D.W.
,
2018
,
Basinal setting and origin of thick (1.8 km) mass-flow dominated Grand Conglomérat diamictites, Kamoa, Democratic Republic of Congo: Resolving climate and tectonic controls during Neoproterozoic glaciations
:
Sedimentology
 , doi: .
Kesler
,
S.E.
,
Appold
,
M.S.
,
Martini
,
A.M.
,
Walter
,
L.M.
,
Huston
,
T.T.
, and
Kyle
,
J.R.
,
1995
,
Na-Cl-Br systematics of mineralizing brines in Mississippi Valley-type deposits
:
Geology
 , v.
23
, p.
641
644
.
Key
,
R.M.
,
Liyungu
,
A.K.
,
Njamu
,
F.M.
,
Somwe
,
V.
,
Banda
,
J.
,
Mosley
,
P.N.
, and
Armstrong
,
R.A.
,
2001
,
The western arm of the Lufilian arc in NW Zambia and its potential for copper mineralization
:
Journal of African Earth Sciences
 , v.
33
, p.
503
528
.
Koziy
,
L.
,
Bull.
,
S.
,
Large
,
R.
, and
Selley
,
D.
,
2009
,
Salt as a fluid driver, and basement as a metal source, for stratiform sediment-hosted copper deposits
:
Geology
 , v.
37
, p.
1107
1110
.
Kukla
,
P.
,
Urai
,
J.
,
Warren
,
J.
,
Reuning
,
L.
,
Becker
,
S.
,
Schoenherr
,
J.
,
Mohr
,
M.
,
Gent
,
H.
,
Abe
,
S.
,
Li
,
S.
,
Desbois
,
G.
,
Schleder
,
Z.
, and
de Keijzer
,
M.
,
2011
,
An integrated, multi-scale approach to salt dynamics and internal dynamics of salt structures
:
American Association of Petroleum Geologists Search and Discovery Article no
 .
40703
,
1
p.
Lefèbvre
,
J.J.
,
1985
,
Le Groupe de Mwashya, mégacyclothème terminal du Roan (Shaba, Zaïre Sud-oriental). II. Volcanisme et dynamique du bassin sédimentaire
:
Musée Royal de l’Afrique Centrale, Tervuren (Belgium), Département de Géologie et Minéralogie, Rapport Annuel 1983–1984
 , p.
121
151
.
Lefèbvre
,
J.J.
, and
Tshiauka
,
T.
,
1986
,
Altérations associées à la mineralisation uranifère de Musoshi (Shaba, Zaire)
:
Tervuren, Belgique, Musée Royal Afrique Centrale Annales, Serie In 8o, Sciences Géologiques
 , no.
92
,
63
p.
Lewis
,
S.
, and
Holness
,
M.
,
1996
,
Equilibrium halite-H2O dihedral angles: High rock-salt permeability in the shallow crust?
:
Geology
 , v.
24
, p.
431
434
.
Loris
,
N.B.T.
,
Chralet
,
J.M.
,
Pechmann
,
E.
,
Clare
,
C.
,
Chabu
,
M.
, and
Quinif
,
Y.
,
1997
,
Caractéristiques minéralogiques, cristallohraphiques, physico-chimiques et âges des minéralisations uranifères de Lwiswishi (Shaba, Zaire)
:
International Cornet Symposium
, September 5–9, 1994,
Mons
,
Royal Academy Overseas Science
, Proceedings, p.
285
306
.
Master
,
S.
,
Rainaud
,
C.
,
Armstrong
,
R.A.
,
Phillips
,
D.
, and
Robb
,
L.J.
,
2003
,
Provenance ages of the Neoproterozoic Katanga Supergroup (Central African Copperbelt), based on SHRIMP U-Pb and laser 40Ar/39Ar dating of detrital zircons and muscovites, with implications for basin evolution
:
University of the Witwatersrand Economic Geology Research Institute Information Circular
 , no.
376
,
33
p.
McCaffrey
,
M.A.
,
Lazar
,
B.
, and
Holland
,
H.D.
,
1987
,
The evaporation path of seawater and the coprecipitation of Br- and K+ with halite
;
Journal of Sedimentary Research
 , v.
57
, p.
928
937
.
McGowan
,
R.R.
,
Roberts
,
S.
,
Foster
,
R.P.
,
Boyce
,
A.J.
, and
Coller
,
D.
,
2003
,
Origin of the copper-cobalt deposits of the Zambian Copperbelt
:
An epigenetic view from Nchanga: Geology
 , v.
31
, p.
497
500
.
Meighan
,
C.J.
,
2015
,
PT Evolution of the Sentinel deposit, northwestern Zambia
: Unpublished M.Sc. thesis,
Golden, Colorado
,
Colorado School of Mines
,
214
p.
Meighan
,
C.J.
,
Kelly
,
N.M.
,
Hitzman
,
M.W.
, and
Wood
,
D.
,
2012
,
Low P-T kyanite growth associated with aluminum metasomatism and quartz-sulfide veining, Sentinel Cu deposit, northwestern Zambia
:
Geological Society of America Abstracts with Programs
 , v.
44
, p.
525
.
Mendelsohn
,
F.
, ed.,
1961
,
The geology of the Northern Rhodesian Copperbelt
:
London, MacDonald
,
523
p.
Meneghel
,
L.
,
1981
,
The occurrence of uranium in the Katanga system of northwestern Zambia
:
Economic Geology
 , v.
76
, p.
56
68
.
Moine
,
B.
,
Guilloux
,
L.
, and
Audeoud
,
D.
,
1986
,
Major element geochemistry of the host rocks in some sediment-hosted copper deposits
, in
Friedrich
,
G.H.
, ed.,
Geology and metallogeny of copper deposits
 :
Berlin
,
Springer-Verlag
, p.
443
460
.
Molak
,
B.
,
1995
,
Some structural and petrological aspects of the Cu (Co) mineralization in the Copperbelt and northwestern provinces of Zambia
:
Royal Museum of Central Africa (Belgium), Annales de la Société Géologique
 , v.
101
, p.
95
102
.
Muchez
,
P.
,
André-Mayer
,
A.S.
,
El Desouky
,
E.L.
, and
Reisberg
,
L.
,
2015
,
Diagenetic origin of the stratiform Cu-Co deposit at Kamoto in the Central African Copperbelt
:
Mineralium Deposita
 , v.
50
, p.
437
447
.
Muchez
,
P.
,
André-Mayer
,
A.S.
,
Dewaele
,
S.
, and
Large
,
R.
,
2017
,
Discussion: “Age of the Zambian Copperbelt”
:
Mineralium Deposita
 , v.
52
, p.
1269
1271
.
Nowecki
,
J.P.
,
2014
,
Tracing seawater evaporation and its role in the formation of sediment-hosted stratiform copper deposits
: Ph.D. thesis,
Southampton, United Kingdom
,
University of Southampton
,
217
p.
Porada
,
H.
, and
Berhorst
,
V.
,
2000
,
Towards a new understanding of the Neoproterozoic-Early Paleozoic Lufilian and Northern Zambezi belts in Zambia and Congo/Zaire
:
Journal of African Earth Sciences
 , v.
30
, p.
727
771
.
Ramm
,
M.
,
1991
,
Porosity-depth trends in reservoir sandstones
:
Theoretical models related to Jurassic sandstones offshore Norway: Marine and Petroleum Geology
 , v.
9
, p.
553
567
.
Ramm
,
M.
, and
Bjørlykke
,
K.
,
1994
,
Porosity/depth trends in reservoir sandstones: Assessing the quantitative effects of varying pore-pressure, temperature history and mineralogy
,
Norwegian shelf data: Clay Minerals
 , v.
29
, p.
475
490
.
Richards
,
J.P.
,
Cumming
,
G.L.
,
Krstic
,
D.
,
Wagner
,
P.A
, and
Spooner
,
E.T.C.
,
1988
,
Pb isotope constraints on the age of sulfide ore deposition and U-Pb age of late uraninite veining at the Musoshi stratiform copper deposit
,
Central African Copperbelt, Zaire: Economic Geology
 , v.
83
, p.
724
741
.
Sarkar
,
A.
,
Nunn
,
J.A.
, and
Hanor
,
J.S.
,
1995
,
Free thermohaline convection beneath allochthonous salt sheets: An agent for salt dissolution and fluid flow in Gulf Coast sediments
:
Journal of Geophysical Research
 , v.
100
, p.
18,085
18,092
.
Schardt
,
C.
,
Cooke
,
D.R.
,
Gemmell
,
J.B.
, and
Large
,
R.R.
,
2001
,
Geochemical modelling of the zoned footwall alteration pipe, Hellyer volcanic-hosted massive sulfide deposit, western Tasmania, Australia
,
Economic Geology
 , v.
96
, p.
1037
1054
.
Schmandt
,
D.
,
Broughton
,
D.
,
Hitzman
,
M.W.
,
Plink-Bjorklund
,
P.
,
Edwards
,
D.
, and
Humphrey
,
J.
,
2013
,
The Kamoa copper deposit, Democratic Republic of Congo
:
Stratigraphy, diagenetic and hydrothermal alteration, and mineralization: Economic Geology
 , v.
108
, p.
1301
1324
.
Schneider
,
J.
,
Melcher
,
F.
, and
Brauns
,
M.
,
2007
,
Concordant ages for the giant Kipushi base metal deposit (DRC) from direct Rb-Sr and Re-Os dating of sulfides
:
Mineralium Deposita
 , v.
42
, p.
791
797
.
Schuh
,
W.
,
Leveille
,
R.A.
,
Fay
,
I.
, and
North
,
R.
,
2012
,
Geology of the Tenke-Fungurume sediment-hosted strata-bound copper-cobalt district, Katanga, Democratic Republic of Congo
:
Society of Economic Geologists Special Publication
 
16
, p.
269
301
.
Selley
,
D.
,
Broughton
,
D.
,
Scott
,
R.
,
Hitzman
,
M.
,
Bull
,
S.
,
Large
,
R.
,
McGoldrick
,
P.
,
Croaker
,
M.
,
Pollington
,
N.
, and
Barra
,
F.
,
2005
,
A new look at the geology of the Zambian Copperbelt: Economic Geology 100th Anniversary Volume
, p.
965
1000
.
Sillitoe
,
R.H.
,
Perelló
,
J.
,
Creaser
,
R.A.
,
Wilton
,
J.
, and
Dawborn
,
T.
,
2015
,
Two ages of copper mineralization in the Mwombezhi dome, northwestern Zambia: Metallogenic implications for the Central African Copperbelt
.
Economic Geology
 , v.
110
, p.
1917
1923
.
Sillitoe
,
R.H.
,
Perelló
,
J.
,
Creaser
,
R.A.
,
Wilton
,
J.
,
Wilson
,
A.
, and
Dawborn
,
T.
,
2017a
,
Age of the Zambian Copperbelt
:
Mineral Deposita
 , v.
52
, p.
1245
1268
.
Sillitoe
,
R.H.
,
Perelló
,
J.
,
Creaser
,
R.A.
,
Wilton
,
J.
,
Wilson
,
A.
, and
Daw-born
,
T.
,
2017b
,
Reply to discussion of “Age of the Zambian Copperbelt” by Hitzman and Broughton and Muchez et al.
:
Mineralium Deposita
 , v.
52
, p.
1277
1281
.
Steven
,
N.
, and
Armstrong
,
R.
,
2003
,
A metamorphosed Proterozoic carbonaceous shale-hosted Co-Ni-Cu deposit at Kalumbila, Kabompo dome
:
The Copperbelt Ore Shale in northwestern Zambia: Economic Geology
 , v.
98
, p.
893
909
.
Stricker
,
S.
,
Jones
,
S.J.
,
Sathar
,
S.
,
Boweb
,
L.
, and
Oxtoby
,
N.
,
2016
,
Exceptional reservoir quality in HPHT reservoir settings: Examples from the Skagerrak Formation of the Heron cluster, North Sea, UK
:
Marine and Petroleum Geology
 , v.
77
, p.
198
215
.
Sweeney
,
M.A.
, and
Binda
,
P.L.
,
1989
,
The role of diagenesis in the formation of the Konkola Cu-Co orebody of the Zambian Copperbelt
:
Geological Association of Canada
 , Special Paper
36
, p.
499
518
.
Swenson
,
J.B.
,
Person
,
M.
,
Raffensperger
,
J.P.
,
Cannon
,
W.F.
,
Woodruff
,
L.G.
, and
Berndt
,
M.E.
,
2004
,
A hydrogeologic model of stratiform copper mineralization in the midcontinent rift system, northern Michigan
,
USA: Geofluids
 , v.
4
, p.
1
22
.
Taylor
,
C.D.
,
Causey
,
J.D.
,
Denning
,
P.D.
,
Hammarstrom
,
J.M.
,
Hayes
,
T.S.
,
Horton
,
J.D.
,
Kirschbaum
,
M.J.
,
Parks
,
H.L.
,
Wilson
,
A.B.
,
Wintzer
,
N.E.
, and
Zientek
,
M.L.
,
2013
,
Descriptive models, grade-tonnage relations, and databases for the assessment of sediment-hosted copper deposits—with emphasis on deposits in the Central African Copperbelt, Democratic Republic of the Congo and Zambia
:
U.S. Geological Survey Scientific Investigations Report 2010–5090–J
 ,
154
p. and data files, http://pubs.usgs.gov/sir/2010/5090/j/.
Thurmond
,
V.L.
,
Potter
,
R.W.
, and
Clynne
,
M.A.
,
1984
,
The densities of saturated solutions of NaCl and KCl from 10o to 105oC
:
U.S. Geological Survey Open-File Report 84–253
 ,
10
p.
Torrealday
,
H.I.
,
Hitzman
,
M.W.
,
Stein
,
H.J.
,
Markley
,
R.J.
,
Armstrong
,
R.
, and
Broughton
,
D.
,
2000
,
Re-Os and U-Pb dating of the vein-hosted mineralization at the Kansanshi copper deposit, northern Zambia
:
Economic Geology
 , v.
95
, p.
1165
1170
.
Trudgill
,
B.
,
2005
,
Salt tectonics short course
:
Golden, Colorado
,
Colorado School of Mines
.
Turlin
,
F.
,
Eglinger
,
A.
,
Vanderhaeghe
,
O.
,
Andre-Mayer
,
A.
,
Poujol
,
M.
,
Bartlett
,
R.
, and
Mercadier
,
J.
,
2016
,
Synmetamorphic Cu remobilization during the Pan-African orogeny
:
Microstructural, petrological and geochronological data on the kyanite-micaschists hosting the Cu (-U) Lumwana deposit in the Western Zambian Copperbelt of the Lufilian belt: Ore Geology Reviews
 , v.
75
, p.
52
75
.
Turner
,
E.C.
,
Broughton
,
D.W.
, and
Brooks
,
T.
,
2018
,
Neoproterozoic carbonate lithofacies and ore distribution at the Kipushi Cu-Zn deposit, Democratic Republic of Congo, and Gayna River Zn camp, Northwest Territories, Canada
:
Economic Geology
 , v.
113
, p.
779
788
.
Vendeville
,
B.C.
, and
Jackson
,
M.P.A.
,
1992
,
The fall of diapirs during thin-skinned extension
:
Marine and Petroleum Geology
 , v.
9
, p.
354
371
.
Walter
,
L.M.
,
Stueber
,
A.M.
, and
Huston
,
T.J.
,
1990
,
Br-Cl-Na systematics in Illinois basin fluids
:
Constraints on fluid origin and evolution: Geology
 , v.
18
, p.
315
318
.
Warren
,
J.
,
1999
,
Evaporites: Their evolution and economics
:
Oxford
,
Blackwell Science
,
438
p.
White
,
D.E.
,
Anderson
,
E.T.
, and
Grubbs
,
D.K.
,
1963
,
Geothermal brine well
:
Mile-deep drill hole may tap ore-bearing magmatic water and rocks undergoing metamorphism: Science
 , v.
139
, p.
919
922
.
Woodhead
,
J.A.
,
2013
,
The Neoproterozoic Roan Group in the Zambian Copperbelt: Sequence stratigraphy, alteration, and mineralization
: Unpublished Ph.D. thesis,
Golden, Colorado, Colorado
School of Mines
,
277
p.
Yardley
,
B.W.D.
, and
Banks
,
D.A.
,
1995
,
The behaviour of chloride and bromide during the metamorphic cycle
, in
Kharaka
,
Y.K.
, and
Chudaev
,
O.V.
, eds.,
Water-rock interaction WRI-8
 :
Rotterdam, Balkema
, p.
625
628
.
Zimba
,
M.
,
2012
,
The composition and role of brines in the formation of copper belt ores
:
Enterprise nickel prospect case study, North-Western Province, Zambia
 : Unpublished M.Sc. thesis,
Leeds, United Kingdom
,
University of Leeds
,
95
p.

Figures & Tables

Fig. 1.

Basic geologic map of the Central African Copperbelt, showing the distribution of various ore types. Abbreviations: CCB = Congolese Copperbelt, DRC = Democratic Republic of Congo, Gp = Group, PGE = platinum group element, Sgp = Subgroup, ZCB = Zambian Copperbelt. Inset shows map location in Africa.

Fig. 1.

Basic geologic map of the Central African Copperbelt, showing the distribution of various ore types. Abbreviations: CCB = Congolese Copperbelt, DRC = Democratic Republic of Congo, Gp = Group, PGE = platinum group element, Sgp = Subgroup, ZCB = Zambian Copperbelt. Inset shows map location in Africa.

Fig. 2.

Stratigraphic framework of the Katangan basin, highlighting thickness contrasts between the central depocenter maximum and peripheral domains where basement inliers are exposed. Generalized locations of the three profiles are shown in Figure 1. The stratigraphic framework for the Roan Group is that of Bull et al. (2011); see text for further discussion and alternative interpretations. Stratal thicknesses depicted in peripheral domains are minima for the respective areas and determined from drilling and analysis of maps. The profile of the depocenter maximum is derived from analysis of Gecamines 1:20,000 maps (noncontiguous maximum thickness measurements for individual formations are shown for middle and upper basin levels), and a gravity-constrained permissible thickness of the nonexposed synrift level of the Lower Roan Subgroup (see Digital App. Fig. A1). A conservative maximum thickness of ~11 km is determined from the balanced cross section shown in Figure 16. Abbreviation: DRC = Democratic Republic of Congo.

Fig. 2.

Stratigraphic framework of the Katangan basin, highlighting thickness contrasts between the central depocenter maximum and peripheral domains where basement inliers are exposed. Generalized locations of the three profiles are shown in Figure 1. The stratigraphic framework for the Roan Group is that of Bull et al. (2011); see text for further discussion and alternative interpretations. Stratal thicknesses depicted in peripheral domains are minima for the respective areas and determined from drilling and analysis of maps. The profile of the depocenter maximum is derived from analysis of Gecamines 1:20,000 maps (noncontiguous maximum thickness measurements for individual formations are shown for middle and upper basin levels), and a gravity-constrained permissible thickness of the nonexposed synrift level of the Lower Roan Subgroup (see Digital App. Fig. A1). A conservative maximum thickness of ~11 km is determined from the balanced cross section shown in Figure 16. Abbreviation: DRC = Democratic Republic of Congo.

Fig. 3.

Summary of published isotopic ages for Cu ± Co, U ± Ni-Co, and Pb-Zn-Cu deposits and prospects of the Central African Copperbelt. Largely reproduced from Sillitoe et al. (2017a), including references therein. Kipushi age data from Schneider et al. (2007). Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt,

Fig. 3.

Summary of published isotopic ages for Cu ± Co, U ± Ni-Co, and Pb-Zn-Cu deposits and prospects of the Central African Copperbelt. Largely reproduced from Sillitoe et al. (2017a), including references therein. Kipushi age data from Schneider et al. (2007). Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt,

Fig. 4.

Geologic map of the Zambian Copperbelt (modified from Selley et al., 2005). With the exception of the Mwashya Subgroup-hosted Frontier deposit, all major deposits are hosted within Lower Roan Subgroup strata. A regional decoupling surface, defined by breccia facies and deformed by upright basement-involved folds, is principally positioned in Upper Roan Subgroup strata. On the western flank of the Kafue anticline, the decoupling surface cuts down-section to middle levels of the Lower Roan Subgroup. Interpolation of the points at which the decoupling surface changes from flat to ramp configuration defines a northwest trace that projects between Chibuluma West and Musoshi and corresponds to the western limits of known Copperbelt Orebody Member strata. Recumbent folds occur above the decoupling zone in the region of the Frontier deposit. Abbreviations: CCB = Congolese Copperbelt, Gp(s) = Group(s), Sgp(s) = Subgroup(s), ZCB = Zambian Copperbelt.

Fig. 4.

Geologic map of the Zambian Copperbelt (modified from Selley et al., 2005). With the exception of the Mwashya Subgroup-hosted Frontier deposit, all major deposits are hosted within Lower Roan Subgroup strata. A regional decoupling surface, defined by breccia facies and deformed by upright basement-involved folds, is principally positioned in Upper Roan Subgroup strata. On the western flank of the Kafue anticline, the decoupling surface cuts down-section to middle levels of the Lower Roan Subgroup. Interpolation of the points at which the decoupling surface changes from flat to ramp configuration defines a northwest trace that projects between Chibuluma West and Musoshi and corresponds to the western limits of known Copperbelt Orebody Member strata. Recumbent folds occur above the decoupling zone in the region of the Frontier deposit. Abbreviations: CCB = Congolese Copperbelt, Gp(s) = Group(s), Sgp(s) = Subgroup(s), ZCB = Zambian Copperbelt.

Fig. 5.

Schematic cross sections illustrating alternative hypotheses for the displacement history of the Zambian Copperbelt decoupling surface. (a) Extensional collapse, with the down-section-cutting ramp on the western flank of the Kafue anticline connecting to a flat positioned at deeper structural levels. (b) Northeast-directed thrusting during the earliest stages of basin inversion, the western ramp representing a short-cut from deeper structural levels. The asymmetry of recumbent folds is consistent with broadly NE-directed thrusting. In both scenarios, localized duplication of the stratigraphy is required. Abbreviations: Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 5.

Schematic cross sections illustrating alternative hypotheses for the displacement history of the Zambian Copperbelt decoupling surface. (a) Extensional collapse, with the down-section-cutting ramp on the western flank of the Kafue anticline connecting to a flat positioned at deeper structural levels. (b) Northeast-directed thrusting during the earliest stages of basin inversion, the western ramp representing a short-cut from deeper structural levels. The asymmetry of recumbent folds is consistent with broadly NE-directed thrusting. In both scenarios, localized duplication of the stratigraphy is required. Abbreviations: Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 6.

Schematic representation of the fold and fault configuration of the Domes region. The Lumwana deposits, Malundwe and Chimiwungo, are positioned along low-angle thrusts, locally refolded by shallowly inclined F2 closures. Enterprise and Sentinel are interpreted to be positioned below a breccia-defined decoupling surface at Upper Roan Subgroup levels. Kansanshi occurs at a supradetachment level, overlying a northward-thickening, synrift package. Major shortening phases were punctuated by episodes of extensionally driven exhumation. Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 6.

Schematic representation of the fold and fault configuration of the Domes region. The Lumwana deposits, Malundwe and Chimiwungo, are positioned along low-angle thrusts, locally refolded by shallowly inclined F2 closures. Enterprise and Sentinel are interpreted to be positioned below a breccia-defined decoupling surface at Upper Roan Subgroup levels. Kansanshi occurs at a supradetachment level, overlying a northward-thickening, synrift package. Major shortening phases were punctuated by episodes of extensionally driven exhumation. Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 7.

Map of the Luina dome region: transition between the Zambian and Congolese Copperbelts (based on A.P. François, unpub. report, 1958; Gecamines 1:20,000 map set). To the south of the dome, the breccia-defined decoupling surface is deformed by high-wavelength, basement-cored anticlines. However, on the northern side of the dome, the decoupling surface forms the sole to relatively low-wavelength, high-amplitude, breccia-cored folds affecting Mines Subgroup and younger strata. The Mines Subgroup-hosted Kimpe deposit is the southernmost of the classic Congolese Copperbelt deposits. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 7.

Map of the Luina dome region: transition between the Zambian and Congolese Copperbelts (based on A.P. François, unpub. report, 1958; Gecamines 1:20,000 map set). To the south of the dome, the breccia-defined decoupling surface is deformed by high-wavelength, basement-cored anticlines. However, on the northern side of the dome, the decoupling surface forms the sole to relatively low-wavelength, high-amplitude, breccia-cored folds affecting Mines Subgroup and younger strata. The Mines Subgroup-hosted Kimpe deposit is the southernmost of the classic Congolese Copperbelt deposits. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 8.

Structural framework of the central Congolese Copperbelt (modified from the Gecamines 1:20,000 map set). See inset for map locations. (a) Distribution of Roan Group strata and their geometric relationship to major faults and folds. A semitransparent mask is overlain on the northern Congolese Copperbelt to highlight features of the central Congolese Copperbelt. In the latter, Roan Group strata are ubiquitously fragmented and define narrow, irregular map patterns that follow the traces of irregular and discontinuous faults. (b) Linear corridors defined by alignment and/or intersections of breccia facies, faults, and fold axial traces. Note that many lineaments crosscut each other with minimal apparent mutual offset. Several of the lineaments project into the northern Congolese Copperbelt domain, where they coincide with major orebodies (e.g., Kakula, Tenke). (c) Gridded bedding dip angles overlain by fold axial traces, breccia facies, and lower levels of the Nguba Group. Systematic geometric and spatial relationships between the grid pattern and fold axial traces reflect inheritance arrays of salt walls and intervening salt withdrawal minibasins. (d) Seismic profile through a diapiric domain, Santos basin, offshore Brazil (from Trudgill, 2005). The arrangement of steeply dipping strata on the flanks of diapirs and low-angle dips within the adjacent minibasins are analogous to the map patterns shown in (c). Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, Congl. = Conglomerate, FZ = fault zone, Gp(s) = Group(s), NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 8.

Structural framework of the central Congolese Copperbelt (modified from the Gecamines 1:20,000 map set). See inset for map locations. (a) Distribution of Roan Group strata and their geometric relationship to major faults and folds. A semitransparent mask is overlain on the northern Congolese Copperbelt to highlight features of the central Congolese Copperbelt. In the latter, Roan Group strata are ubiquitously fragmented and define narrow, irregular map patterns that follow the traces of irregular and discontinuous faults. (b) Linear corridors defined by alignment and/or intersections of breccia facies, faults, and fold axial traces. Note that many lineaments crosscut each other with minimal apparent mutual offset. Several of the lineaments project into the northern Congolese Copperbelt domain, where they coincide with major orebodies (e.g., Kakula, Tenke). (c) Gridded bedding dip angles overlain by fold axial traces, breccia facies, and lower levels of the Nguba Group. Systematic geometric and spatial relationships between the grid pattern and fold axial traces reflect inheritance arrays of salt walls and intervening salt withdrawal minibasins. (d) Seismic profile through a diapiric domain, Santos basin, offshore Brazil (from Trudgill, 2005). The arrangement of steeply dipping strata on the flanks of diapirs and low-angle dips within the adjacent minibasins are analogous to the map patterns shown in (c). Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, Congl. = Conglomerate, FZ = fault zone, Gp(s) = Group(s), NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 9.

Aeromagnetic data map (total magnetic intensity, reduced to the pole, first vertical derivative) of the Domes region and its transition into the southern Congolese Copperbelt. The boundary, shown as curvilinear, segmented, predominantly NNE-dipping fault zone, is defined in part by the abrupt northward increase in the areal extent of Roan Group breccia complexes and more complex noncylindrical fold patterns. In this region, breccia facies has a distinctive high-frequency/amplitude magnetic texture, due to significant volumes of dismembered mafic intrusives. Map location shown in inset. Abbreviations: CCCB = central Congolese Copperbelt, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 9.

Aeromagnetic data map (total magnetic intensity, reduced to the pole, first vertical derivative) of the Domes region and its transition into the southern Congolese Copperbelt. The boundary, shown as curvilinear, segmented, predominantly NNE-dipping fault zone, is defined in part by the abrupt northward increase in the areal extent of Roan Group breccia complexes and more complex noncylindrical fold patterns. In this region, breccia facies has a distinctive high-frequency/amplitude magnetic texture, due to significant volumes of dismembered mafic intrusives. Map location shown in inset. Abbreviations: CCCB = central Congolese Copperbelt, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 10.

Cross section through the Kambove deposit (refer to Fig. 8 for location), showing the typical chaotic structural geometry of the Mines Subgroup-hosted ores of the central Congolese Copperbelt (after Cailteux, 1986). The breccia complex and its carapace of Mwashya Subgroup strata have been emplaced to the level of the Ngule Subgroup. The level of ore is highlighted in the thin red traces, demonstrating preservation of stratabound form, despite the structural complexity. Abbreviations: Gp = Group, undiff. = undifferentiated.

Fig. 10.

Cross section through the Kambove deposit (refer to Fig. 8 for location), showing the typical chaotic structural geometry of the Mines Subgroup-hosted ores of the central Congolese Copperbelt (after Cailteux, 1986). The breccia complex and its carapace of Mwashya Subgroup strata have been emplaced to the level of the Ngule Subgroup. The level of ore is highlighted in the thin red traces, demonstrating preservation of stratabound form, despite the structural complexity. Abbreviations: Gp = Group, undiff. = undifferentiated.

Fig. 11.

Map of the Kipese-Shinkolobwe region, central Congolese Copperbelt, highlighting the complex configuration of brecciated Roan Group strata and thickness variation at Nguba Group levels, the latter most evident in the Tantara valley. The map patterns represent a rarely exposed profile through a minibasin, with thinning of Grand Conglomerate strata in particular onto the flanks of the former Kipese diapir. Note also the transgressive nature of the breccia facies, cutting upsection to Kundelungu Group stratigraphic levels. Refer to Digital Appendix Table A2 for a detailed stratigraphic framework for the Nguba and Kundelungu Groups. Modified from the Gecamines 1:20,000 map set. Abbreviations: CCCB = central Congolese Copperbelt, Congl. = Conglomerate, Gp = Group, NCCB = northern Congolese Copperbelt, Sgp = Subgroup, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 11.

Map of the Kipese-Shinkolobwe region, central Congolese Copperbelt, highlighting the complex configuration of brecciated Roan Group strata and thickness variation at Nguba Group levels, the latter most evident in the Tantara valley. The map patterns represent a rarely exposed profile through a minibasin, with thinning of Grand Conglomerate strata in particular onto the flanks of the former Kipese diapir. Note also the transgressive nature of the breccia facies, cutting upsection to Kundelungu Group stratigraphic levels. Refer to Digital Appendix Table A2 for a detailed stratigraphic framework for the Nguba and Kundelungu Groups. Modified from the Gecamines 1:20,000 map set. Abbreviations: CCCB = central Congolese Copperbelt, Congl. = Conglomerate, Gp = Group, NCCB = northern Congolese Copperbelt, Sgp = Subgroup, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 12.

Isopachs of Nguba Group strata, overlain by macroscopic structural elements, Roan Group map patterns, and major deposits. Thickness data were derived from outcrop traverses indicated on the Gecamines 1:20,000 map set. Tabulated data and the method of thickness calculation are given in Digital Appendix Table A3 and Figure A3, respectively. Classification of data is based on the quality of parameters used to calculate unit thickness: i.e., unit outcrop, boundary constraints, and bedding orientation. Those data labeled “robust” were determined from traverses where these parameters are moderately to well constrained, with measurements of sufficient precision to examine macroscale thickness variation. Those data labeled “minimum” thickness relate to measurements made where upper or lower unit boundaries are faulted or do not project to surface. “Maximum” thickness refers to data where bedding orientation is poorly constrained, and the calculation assumes a vertical dip. Those data labeled “approximate” lack sufficient outcrop, formation boundary constraints, and/or bedding orientation to derive an accurate thickness measurement. All data are used in construction of the grids, as nonrobust classes typically vary systematically with robust data, and while they do not represent absolute thicknesses, they contribute to important macroscale geometric patterns; care should be taken where nonrobust data coincide with local anomalies in the grids. (a) Upper Nguba Group patterns reveal abrupt southward thickening across major ENE-, NW-, and WNW-striking fault zones. Several of the fault zones coincide with small elongate domains of anomalously condensed strata (red arrows). These represent areas of low-accommodation development on the crests of salt walls (cf. Fig. 8d). Classic stratiform Congolese Copperbelt ores occur mainly in condensed parts of the basin, those close to the Kakanda-Luisha fault zone along a major subbasin edge, and the larger northern deposits within a thin platform. (b) Grand Conglomerate patterns have an opposing relationship, with strata apparently thinner in the southern part of the basin system. This is an artefact of sample distribution, with most data occurring along the crests of Roan Group breccia-cored antiforms and breccia-lined fault zones. The restriction of data to salt wall crests means that the true volume of Grand Conglomerate is greatly underrepresented. The positive thickness anomaly at the fringe of the Kipese breccia to the south is more representative and corresponds to the minibasin shown in Figure 11. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 12.

Isopachs of Nguba Group strata, overlain by macroscopic structural elements, Roan Group map patterns, and major deposits. Thickness data were derived from outcrop traverses indicated on the Gecamines 1:20,000 map set. Tabulated data and the method of thickness calculation are given in Digital Appendix Table A3 and Figure A3, respectively. Classification of data is based on the quality of parameters used to calculate unit thickness: i.e., unit outcrop, boundary constraints, and bedding orientation. Those data labeled “robust” were determined from traverses where these parameters are moderately to well constrained, with measurements of sufficient precision to examine macroscale thickness variation. Those data labeled “minimum” thickness relate to measurements made where upper or lower unit boundaries are faulted or do not project to surface. “Maximum” thickness refers to data where bedding orientation is poorly constrained, and the calculation assumes a vertical dip. Those data labeled “approximate” lack sufficient outcrop, formation boundary constraints, and/or bedding orientation to derive an accurate thickness measurement. All data are used in construction of the grids, as nonrobust classes typically vary systematically with robust data, and while they do not represent absolute thicknesses, they contribute to important macroscale geometric patterns; care should be taken where nonrobust data coincide with local anomalies in the grids. (a) Upper Nguba Group patterns reveal abrupt southward thickening across major ENE-, NW-, and WNW-striking fault zones. Several of the fault zones coincide with small elongate domains of anomalously condensed strata (red arrows). These represent areas of low-accommodation development on the crests of salt walls (cf. Fig. 8d). Classic stratiform Congolese Copperbelt ores occur mainly in condensed parts of the basin, those close to the Kakanda-Luisha fault zone along a major subbasin edge, and the larger northern deposits within a thin platform. (b) Grand Conglomerate patterns have an opposing relationship, with strata apparently thinner in the southern part of the basin system. This is an artefact of sample distribution, with most data occurring along the crests of Roan Group breccia-cored antiforms and breccia-lined fault zones. The restriction of data to salt wall crests means that the true volume of Grand Conglomerate is greatly underrepresented. The positive thickness anomaly at the fringe of the Kipese breccia to the south is more representative and corresponds to the minibasin shown in Figure 11. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 13.

Structure and distribution of major Cu ± Co ores of the northern Congolese Copperbelt; semitransparent overlay masks features of the central Congolese Copperbelt. (a) Map of the northern Congolese Copperbelt, modified from the Gecamines 1:20,000 map set. The region is divided into four domains on the basis of structural style: Kamoa, Kolwezi, Pumpi, and Tenke domains. (b) Cross section B-B’ of the northern and central Congolese Copperbelt, highlighting the contrasting thin- and thick-skinned structural styles, respectively. See text for discussion. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, Congl. = Conglomerate, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, RSL = relative sea level, SCCB = southern Congolese Copperbelt, Sgp(s) = Subgroup(s), ZCB = Zambian Copperbelt.

Fig. 13.

Structure and distribution of major Cu ± Co ores of the northern Congolese Copperbelt; semitransparent overlay masks features of the central Congolese Copperbelt. (a) Map of the northern Congolese Copperbelt, modified from the Gecamines 1:20,000 map set. The region is divided into four domains on the basis of structural style: Kamoa, Kolwezi, Pumpi, and Tenke domains. (b) Cross section B-B’ of the northern and central Congolese Copperbelt, highlighting the contrasting thin- and thick-skinned structural styles, respectively. See text for discussion. Abbreviations: CCB = Congolese Copperbelt, CCCB = central Congolese Copperbelt, Congl. = Conglomerate, FZ = fault zone, Gp = Group, NCCB = northern Congolese Copperbelt, RSL = relative sea level, SCCB = southern Congolese Copperbelt, Sgp(s) = Subgroup(s), ZCB = Zambian Copperbelt.

Fig. 14.

Internal structural geometry of the Kolwezi klippe, modified from A.P. François, (unpub. report, 1973). (a) Surface map illustrating preservation of complexly thrusted and folded Mines and lower Dipeta Subgroups strata. The arenaceous stratigraphic footwall to the Mines Subgroup, R.A.T, is unusually coherent and lacks significant volumes of breccia facies. Refer to Digital Appendix Table A2 for lithologic descriptions of units. (b) Klippe profile showing a downward-converging, principally low angle fault array and associated recumbent to inclined folds, producing a low-amplitude antiformal stack centered near the broad synformal closure defined by the form of the sole thrust. Note that some of the faults appear to have accommodated net loss of stratigraphy, a feature interpreted to indicate an extensional history prior to inversion. Highest Cu grades and tonnages occur within the central part of the profile. (c) Restored profile interpreting downward-converging listric normal faults that originally rooted into an attenuated salt layer, now defined by a metric-scale breccia body lining the surface of the sole thrust. The principal resource is positioned at flip in the polarity of the listric faults. The correspondence between metal endowment and structural configuration is considered to in part, at least, reflect fluid focusing through the lower zone of fault convergence. The present configuration records ~50% north-south shortening. Abbreviations: CCCB = central Congolese Copperbelt, C.M.N. = Calcaire à Minerais Noir, Gp = Group, NCCB = northern Congolese Copperbelt, R.A.T. = Roches Argilo-Talqueuses, SCCB = southern Congolese Copperbelt, S.D. = Schistes Dolomitiques, undiff. = undifferentiated, ZCB = Zambian Copperbelt.

Fig. 14.

Internal structural geometry of the Kolwezi klippe, modified from A.P. François, (unpub. report, 1973). (a) Surface map illustrating preservation of complexly thrusted and folded Mines and lower Dipeta Subgroups strata. The arenaceous stratigraphic footwall to the Mines Subgroup, R.A.T, is unusually coherent and lacks significant volumes of breccia facies. Refer to Digital Appendix Table A2 for lithologic descriptions of units. (b) Klippe profile showing a downward-converging, principally low angle fault array and associated recumbent to inclined folds, producing a low-amplitude antiformal stack centered near the broad synformal closure defined by the form of the sole thrust. Note that some of the faults appear to have accommodated net loss of stratigraphy, a feature interpreted to indicate an extensional history prior to inversion. Highest Cu grades and tonnages occur within the central part of the profile. (c) Restored profile interpreting downward-converging listric normal faults that originally rooted into an attenuated salt layer, now defined by a metric-scale breccia body lining the surface of the sole thrust. The principal resource is positioned at flip in the polarity of the listric faults. The correspondence between metal endowment and structural configuration is considered to in part, at least, reflect fluid focusing through the lower zone of fault convergence. The present configuration records ~50% north-south shortening. Abbreviations: CCCB = central Congolese Copperbelt, C.M.N. = Calcaire à Minerais Noir, Gp = Group, NCCB = northern Congolese Copperbelt, R.A.T. = Roches Argilo-Talqueuses, SCCB = southern Congolese Copperbelt, S.D. = Schistes Dolomitiques, undiff. = undifferentiated, ZCB = Zambian Copperbelt.

Fig. 15.

Geometry of the Mongo structure (see also Fig. 13 for location). (a) Gridded bedding dip angles from the northern Pumpi domain, illustrating concentration of rotational strain within the Mongo structure. Neighboring areas of Kundelungu Group strata have contrastingly low bedding dip magnitudes. (b) Palinspastic reconstruction of the Mongo structure, revealing thickened Ngule Subgroup strata within a crestal graben, i.e., collapsed diapiric head. During early Ngule Subgroup deposition, salt encompassed rafts of the Nguba and lower Kundelungu strata; a salt stock canopy was emplaced at the level of the lower Ngule Subgroup. Subsequent diapiric collapse is interpreted to have been driven by extensional separation of the neighboring rafts. Abbreviations: CCCB = central Congolese Copperbelt, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 15.

Geometry of the Mongo structure (see also Fig. 13 for location). (a) Gridded bedding dip angles from the northern Pumpi domain, illustrating concentration of rotational strain within the Mongo structure. Neighboring areas of Kundelungu Group strata have contrastingly low bedding dip magnitudes. (b) Palinspastic reconstruction of the Mongo structure, revealing thickened Ngule Subgroup strata within a crestal graben, i.e., collapsed diapiric head. During early Ngule Subgroup deposition, salt encompassed rafts of the Nguba and lower Kundelungu strata; a salt stock canopy was emplaced at the level of the lower Ngule Subgroup. Subsequent diapiric collapse is interpreted to have been driven by extensional separation of the neighboring rafts. Abbreviations: CCCB = central Congolese Copperbelt, Gp = Group, NCCB = northern Congolese Copperbelt, SCCB = southern Congolese Copperbelt, ZCB = Zambian Copperbelt.

Fig. 16.

Balanced cross section of the northern and central Congolese Copperbelt through the Tenke domain, looking west. Refer to Figure 13 for the section trace. (a) The present configuration involves ~35% orogenic shortening (or a stretch of 0.65), with <4-km displacement on the northern thrust front. Major deposits are projected onto section. Those positioned south of the Kansuki fault zone occupy the cores of the pinched diapirs. The Tenke deposits, exposed in a tectonic window, are shown to occur below a carapace of relatively condensed Nguba and Kundelungu Groups strata, now largely lost to erosion. (b) The restored profile reveals a southward-thickening series of minibasins, bounded above the Lower Roan Subgroup synrift package by an array of high-amplitude salt walls (i.e., Mines and Dipeta Subgroups and salt-related breccia). Minibasins are interpreted to have welded at the bases of their thickened cores to the synrift package. Hosts to classic Cu-Co ores are considered to have originated close to salt welds and were repositioned to higher structural levels via salt flow and/or orogenic thickening of the pile. (c) Expanded views of the Tenke domain with schematic representation of Mines Subgroup écailles. The restored profile shows the Kansuki fault zone as a diapiric structure positioned above a major synrift subbasin boundary and between basal-welded minibasins. The northern minibasin, about which known high-grade Cu mineralization occurs, corresponds to the northern limit of Mines Subgroup écailles (see also Fig. 13). The present configuration shows inflation and internal buckling of the Tenke upper Roan Group sequence due to northward impingement of the southern minibasin and partial underthrusting of the Kansuki fault zone. A Grand Conglomerate-cored syncline is the only preserved remnant of the Nguba-Kundelungu Groups carapace in the region of the Tenke deposits and is interpreted to represent the keel of an inverted minibasin. Abbreviations: CCB = Congolese Copperbelt, FZ = fault zone, Gp = Group, Kund. = Kundelungu, PGE = platinum group element, Sgp(s) = Subgroup(s).

Fig. 16.

Balanced cross section of the northern and central Congolese Copperbelt through the Tenke domain, looking west. Refer to Figure 13 for the section trace. (a) The present configuration involves ~35% orogenic shortening (or a stretch of 0.65), with <4-km displacement on the northern thrust front. Major deposits are projected onto section. Those positioned south of the Kansuki fault zone occupy the cores of the pinched diapirs. The Tenke deposits, exposed in a tectonic window, are shown to occur below a carapace of relatively condensed Nguba and Kundelungu Groups strata, now largely lost to erosion. (b) The restored profile reveals a southward-thickening series of minibasins, bounded above the Lower Roan Subgroup synrift package by an array of high-amplitude salt walls (i.e., Mines and Dipeta Subgroups and salt-related breccia). Minibasins are interpreted to have welded at the bases of their thickened cores to the synrift package. Hosts to classic Cu-Co ores are considered to have originated close to salt welds and were repositioned to higher structural levels via salt flow and/or orogenic thickening of the pile. (c) Expanded views of the Tenke domain with schematic representation of Mines Subgroup écailles. The restored profile shows the Kansuki fault zone as a diapiric structure positioned above a major synrift subbasin boundary and between basal-welded minibasins. The northern minibasin, about which known high-grade Cu mineralization occurs, corresponds to the northern limit of Mines Subgroup écailles (see also Fig. 13). The present configuration shows inflation and internal buckling of the Tenke upper Roan Group sequence due to northward impingement of the southern minibasin and partial underthrusting of the Kansuki fault zone. A Grand Conglomerate-cored syncline is the only preserved remnant of the Nguba-Kundelungu Groups carapace in the region of the Tenke deposits and is interpreted to represent the keel of an inverted minibasin. Abbreviations: CCB = Congolese Copperbelt, FZ = fault zone, Gp = Group, Kund. = Kundelungu, PGE = platinum group element, Sgp(s) = Subgroup(s).

Fig. 17.

Cross section of the Kwatebala deposit, easternmost of the Tenke deposits (refer to Fig. 13a for deposit location; modified from François, 1986). The geometry involves an early stage thrust stack, refolded by tight upright closures and open recumbent closures. Ore occurs within the Kamoto dolomite and lowest levels of the S.D.-C.M.N. sequence, and in this case, only the structurally highest thrust sheet contains economic mineralization. It is interpreted that part of the early thrust history relates to localized shortening at the margins of downward-impinging, salt withdrawal basins. Abbreviations: C.M.N. = Calcaire à Minerais Noir, S.D. = Schistes Dolomitiques, Sgp = Subgroup.

Fig. 17.

Cross section of the Kwatebala deposit, easternmost of the Tenke deposits (refer to Fig. 13a for deposit location; modified from François, 1986). The geometry involves an early stage thrust stack, refolded by tight upright closures and open recumbent closures. Ore occurs within the Kamoto dolomite and lowest levels of the S.D.-C.M.N. sequence, and in this case, only the structurally highest thrust sheet contains economic mineralization. It is interpreted that part of the early thrust history relates to localized shortening at the margins of downward-impinging, salt withdrawal basins. Abbreviations: C.M.N. = Calcaire à Minerais Noir, S.D. = Schistes Dolomitiques, Sgp = Subgroup.

Fig. 18.

Schematic model for the emplacement of the Kolwezi klippe (looking west): present configuration, and profile restored to the level of the upper Ngule Subgroup. Approximate position of section indicated in Figure 13. The presently preserved antiformal stack of Mines Subgroup écailles is interpreted to have been positioned at the base of rafted minibasin, the latter now lost to erosion. The highest-grade core of the klippe is shown to have been symmetrically arranged about a salt weld prior to shortening. During the final shortening phase, bounding rafts converged to weld laterally, causing the klippe to be emplaced vertically to the level of an Ngule Subgroup salt stock canopy. Abbreviation: Sgp = Subgroup.

Fig. 18.

Schematic model for the emplacement of the Kolwezi klippe (looking west): present configuration, and profile restored to the level of the upper Ngule Subgroup. Approximate position of section indicated in Figure 13. The presently preserved antiformal stack of Mines Subgroup écailles is interpreted to have been positioned at the base of rafted minibasin, the latter now lost to erosion. The highest-grade core of the klippe is shown to have been symmetrically arranged about a salt weld prior to shortening. During the final shortening phase, bounding rafts converged to weld laterally, causing the klippe to be emplaced vertically to the level of an Ngule Subgroup salt stock canopy. Abbreviation: Sgp = Subgroup.

Fig. 19.

Schematic model for the evolution of the Katangan basin through orogenesis. The profiles look broadly westward. (a) Basin form immediately prior to orogenesis, involving crude symmetrical tapering about a central depocenter maximum. Diapirs punctuate upper levels of the basin fill in the Congolese Copperbelt, reaching maximum amplitude in central and southern parts. Thin salt in the Domes region and Zambian Copperbelt remained essentially unmodified. The northern basin margin, which records onset of growth during Mwashya Subgroup deposition (marginal coarse-grained siliciclastics), was devoid of salt. (b) D1 shortening, involving burial and high-grade metamorphism of the Domes region, and basin-wide decoupling at the level of salt. Diapirs were squeezed in the Congolese Copperbelt to become fully emergent, linking at the surface in northern parts to produce a salt stock canopy at lower levels of the Ngule Subgroup. Emplacement of a salt allochthon at the level of the Grand Conglomerate at the northern edge of the basin. (c) Extensionally driven decompression in the Domes region and extensional collapse along the evaporitic décollement to produce crestal grabens in the northern Congolese Copperbelt. (d) D2 shortening, involving basin-wide foliation development. Thick-skinned, basement-involved folding of the Zambian Copperbelt, low-angle thrusting in the Domes region partly associated with southward exhumation of the depocenter maximum, continued emergence of diapirs, and vertical stacking of rafts in the northern Congolese Copperbelt. Distribution of deposit types shown with no temporal significance implied. See text for discussion. Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 19.

Schematic model for the evolution of the Katangan basin through orogenesis. The profiles look broadly westward. (a) Basin form immediately prior to orogenesis, involving crude symmetrical tapering about a central depocenter maximum. Diapirs punctuate upper levels of the basin fill in the Congolese Copperbelt, reaching maximum amplitude in central and southern parts. Thin salt in the Domes region and Zambian Copperbelt remained essentially unmodified. The northern basin margin, which records onset of growth during Mwashya Subgroup deposition (marginal coarse-grained siliciclastics), was devoid of salt. (b) D1 shortening, involving burial and high-grade metamorphism of the Domes region, and basin-wide decoupling at the level of salt. Diapirs were squeezed in the Congolese Copperbelt to become fully emergent, linking at the surface in northern parts to produce a salt stock canopy at lower levels of the Ngule Subgroup. Emplacement of a salt allochthon at the level of the Grand Conglomerate at the northern edge of the basin. (c) Extensionally driven decompression in the Domes region and extensional collapse along the evaporitic décollement to produce crestal grabens in the northern Congolese Copperbelt. (d) D2 shortening, involving basin-wide foliation development. Thick-skinned, basement-involved folding of the Zambian Copperbelt, low-angle thrusting in the Domes region partly associated with southward exhumation of the depocenter maximum, continued emergence of diapirs, and vertical stacking of rafts in the northern Congolese Copperbelt. Distribution of deposit types shown with no temporal significance implied. See text for discussion. Abbreviations: CCB = Congolese Copperbelt, Gp = Group, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 20.

Na-Cl-Br variations in leachate data from various Central African Copperbelt deposit types and additional data from the Kupferschiefer. Cl/Br and Na/Br are molar ratios. (a) Previously published Central African Copperbelt data: Kipushi data sourced from Heijlen et al. (2008), the remainder from Nowecki (2014). (b) Newly presented data from the Congolese Copperbelt, Zambian Copperbelt, Domes region, and Kupferschiefer. (c) Expanded view of (b) highlighting positions of data relative to the seawater evaporation trend. Abbreviations: CCB = Congolese Copperbelt, NaCli = initial halite precipitation, NaClf = final halite precipitation, PGE = platinum group element, SET = seawater evaporation trend, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 20.

Na-Cl-Br variations in leachate data from various Central African Copperbelt deposit types and additional data from the Kupferschiefer. Cl/Br and Na/Br are molar ratios. (a) Previously published Central African Copperbelt data: Kipushi data sourced from Heijlen et al. (2008), the remainder from Nowecki (2014). (b) Newly presented data from the Congolese Copperbelt, Zambian Copperbelt, Domes region, and Kupferschiefer. (c) Expanded view of (b) highlighting positions of data relative to the seawater evaporation trend. Abbreviations: CCB = Congolese Copperbelt, NaCli = initial halite precipitation, NaClf = final halite precipitation, PGE = platinum group element, SET = seawater evaporation trend, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 21.

Hydrologic model for the Katangan basin, showing permissible timings for the development of various alteration assemblages and types of mineralization, relative to stages of basin evolution. (a) First rift stage: deposition of Lower and Upper Roan Subgroups. Deposition of basin-wide salt sheets resulted in recharge of a basal arenite-dominated aquifer by an oxidizing Mg-rich residual evaporitic brine. Replacement of detrital phases by Mg chlorite in the depocenter maximum had the potential to initiate repartitioning of alkali elements (± light rare earth elements, Ni) toward basin-peripheral parts of the aquifer system. Possible early stage stratiform Cu-Co mineralization would have been restricted to the outer fringes of the salt sheets, where reducing elements, such as the Copperbelt Orebody Member in the Zambian Copperbelt, were in direct hydrological communication with the aquifer. (b) Mature stage of basin growth: deposition of the Nguba Group. Invigorated fluid flow in the basal aquifer was driven by increasing thermal contrast between the progressively subsiding depocenter maximum and the basin margins, contributing to the regional alteration zonation developed at subsalt levels. Primary hydrologic connectivity between the aquifer and reducing elements at the basin margins (including the Grand Conglomerate) demands that at least some component of the stratiform Cu (Co) ores would have formed at this time. Onset of halokinesis during Mwashya Subgroup and/or Grand Conglomerate deposition led to fragmentation of intrasalt strata and the potential for breaching of the basal salt seal at sites of welding. In basin marginal positions, where salt was originally thin, convective fluid cells had the potential to infiltrate intrasalt strata (i.e., multiple points of contact between intrasalt strata and the basal aquifer). This hydrologic framework is considered to have been fundamental in the formation of laterally extensive classic stratiform Congolese Copperbelt ores. The onset of salt dissolution potentially occurred during this period, with emergence of diapirs at the surface, or interaction of suprasalt, halite-undersaturated pores with domains of abnormally permeable salt; both of these conditions were most likely met in the depocenter maximum, where salt was originally thickest and most deeply buried. The hydrologic architecture depicted in center of the profile, involving connectivity of the basal aquifer to surface via a combination of welding and diapir collapse, may account for the mixed residual brine-halite dissolution-related brine Na-Cl-Br signature of Shinkolobwe solute data. (c) Orogenesis provided the strongest driving force for fluid flow in the basin’s history (Sillitoe et al., 2017a). At the initiation of basin inversion (noting that this may not have been synchronous at the basin scale), the fundamental hydrologic framework at the basin margins was likely little modified from early and/or intermediate stages of basin growth, allowing for formation and/or upgrading of existing, stratiform ores; however, effective fluid flow into the halokinetically dismembered intrasalt strata was, at this point, unlikely, with the possible exception of basin marginal zones where salt was originally very thin. As tectonism progressed and parts of the basin were deeply buried and metamorphosed salt dissolution pervaded the middle levels of the structural pile, leading to a fundamental reorganization of basin hydrology. Fluids derived from a mixture of residual- and salt-dissolution–related brine were directed along highly permeable former salt layers and diapirs to interact with suprasalt-reducing elements, forming fracture-controlled ore types. Partitioning of Na-dominant alteration phases largely within and above the level of dissolving salt is interpreted to reflect relatively high temperature fluid flow within the core of the orogen. Abbreviations: CCB = Congolese Copperbelt, COM = Copperbelt Orebody Member, Congl. = Conglomerate, Gp = Group, PGE = platinum group element, REE= rare earth element, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Fig. 21.

Hydrologic model for the Katangan basin, showing permissible timings for the development of various alteration assemblages and types of mineralization, relative to stages of basin evolution. (a) First rift stage: deposition of Lower and Upper Roan Subgroups. Deposition of basin-wide salt sheets resulted in recharge of a basal arenite-dominated aquifer by an oxidizing Mg-rich residual evaporitic brine. Replacement of detrital phases by Mg chlorite in the depocenter maximum had the potential to initiate repartitioning of alkali elements (± light rare earth elements, Ni) toward basin-peripheral parts of the aquifer system. Possible early stage stratiform Cu-Co mineralization would have been restricted to the outer fringes of the salt sheets, where reducing elements, such as the Copperbelt Orebody Member in the Zambian Copperbelt, were in direct hydrological communication with the aquifer. (b) Mature stage of basin growth: deposition of the Nguba Group. Invigorated fluid flow in the basal aquifer was driven by increasing thermal contrast between the progressively subsiding depocenter maximum and the basin margins, contributing to the regional alteration zonation developed at subsalt levels. Primary hydrologic connectivity between the aquifer and reducing elements at the basin margins (including the Grand Conglomerate) demands that at least some component of the stratiform Cu (Co) ores would have formed at this time. Onset of halokinesis during Mwashya Subgroup and/or Grand Conglomerate deposition led to fragmentation of intrasalt strata and the potential for breaching of the basal salt seal at sites of welding. In basin marginal positions, where salt was originally thin, convective fluid cells had the potential to infiltrate intrasalt strata (i.e., multiple points of contact between intrasalt strata and the basal aquifer). This hydrologic framework is considered to have been fundamental in the formation of laterally extensive classic stratiform Congolese Copperbelt ores. The onset of salt dissolution potentially occurred during this period, with emergence of diapirs at the surface, or interaction of suprasalt, halite-undersaturated pores with domains of abnormally permeable salt; both of these conditions were most likely met in the depocenter maximum, where salt was originally thickest and most deeply buried. The hydrologic architecture depicted in center of the profile, involving connectivity of the basal aquifer to surface via a combination of welding and diapir collapse, may account for the mixed residual brine-halite dissolution-related brine Na-Cl-Br signature of Shinkolobwe solute data. (c) Orogenesis provided the strongest driving force for fluid flow in the basin’s history (Sillitoe et al., 2017a). At the initiation of basin inversion (noting that this may not have been synchronous at the basin scale), the fundamental hydrologic framework at the basin margins was likely little modified from early and/or intermediate stages of basin growth, allowing for formation and/or upgrading of existing, stratiform ores; however, effective fluid flow into the halokinetically dismembered intrasalt strata was, at this point, unlikely, with the possible exception of basin marginal zones where salt was originally very thin. As tectonism progressed and parts of the basin were deeply buried and metamorphosed salt dissolution pervaded the middle levels of the structural pile, leading to a fundamental reorganization of basin hydrology. Fluids derived from a mixture of residual- and salt-dissolution–related brine were directed along highly permeable former salt layers and diapirs to interact with suprasalt-reducing elements, forming fracture-controlled ore types. Partitioning of Na-dominant alteration phases largely within and above the level of dissolving salt is interpreted to reflect relatively high temperature fluid flow within the core of the orogen. Abbreviations: CCB = Congolese Copperbelt, COM = Copperbelt Orebody Member, Congl. = Conglomerate, Gp = Group, PGE = platinum group element, REE= rare earth element, Sgp = Subgroup, ZCB = Zambian Copperbelt.

Table 1.

Compositions of Leachates from Various Central African Copperbelt Deposit Types and Kupferschiefer Ore

SampleRegionDeposit typeDepositMineralNa (ng)NH4 (ng)K (ng)Mg (ng)Ca (ng)Sr (ng)Ba (ng)Cl (ng)Br (ng)SO4 (ng)Cat/AnCl/BrNa/Br
NS137 cpyZCBClassic arenite-hosted Cu-CoChibuluma WestCcp-Crl-Dol81.105.118.735.4717.46ndnd74.920.81204.000.83209349
NS137 1281 cpy/pyZCBClassic arenite-hosted Cu-CoChibuluma WestPy-Ccp139.775.1448.254.09185.910.622.01286.061.02457.590.98631475
NS137 1300.5ZCBClassic arenite-hosted Cu-CoChibuluma WestCcp-Crl108.841.8210.072.441.62ndnd111.001.1584.191.08217328
NC-car (NW)ZCBClassic arenite-hosted Cu-CoNchangaCrl369.8116.71176.371.7829.410.673.93875.794.73232.140.77417272
KN18 1163.3ZCBClassic argillite-hosted Cu-CoKonkola NorthDol272.431.9571.41106.31195.254.562.81591.993.4982.850.66383272
Mindola dupZCBClassic argillite-hosted Cu-CoMindolaDol-Anh121.731.16104.2783.95373.821.65nd209.561.64536.801.09289259
E1099BX14CCBClassic Cu-CoEtoileCct (in vein)348.012.3117.601.741.85nd1.98500.891.0964.171.031,0331,107
KYA219cpyCCBClassic Cu-CoKamoyaCcp209.216.83198.7440.89473.5119.8433.661,548.9812.94400.110.8027056
H3410 carCCBClassic Cu-CoKisanfuCrl217.365.27374.7633.30195.6813.0439.311,421.3811.220.000.6728667
KOV558/294.6CCBClassic Cu-CoKolweziCrl104.710.6742.1740.08198.614.7520.18673.555.6953.410.7926764
Luilu22 213CCBClassic Cu-CoKolweziCrl-Bn88.301.2053.8373.9570.855.4136.78416.982.3939.931.22394129
KTO559 421.5CCBClassic Cu-CoKolweziCrl-Bn48.790.1733.760.7838.171.770.65248.841.4413.100.68390118
KTO565 543.9CCBClassic Cu-CoKolweziCct136.170.2651.250.15104.816.597.38861.396.731.910.5228870
KTO565 553.9CCBClassic Cu-CoKolweziMixed sulfide98.541.05111.5539.30458.1611.2462.24748.024.27261.650.6239580
KTO565 548CCBClassic Cu-CoKolweziMixed sulfide412.971.9864.3432.19130.143.284.88996.512.9655.600.60758485
KTO565 543.9 dol/sulCCBClassic Cu-CoKolweziDol-Cct-Crl96.080.84125.8584.59420.834.0751.29767.465.7131.821.3330359
KADI19 165.9CCBClassic Cu-CoKolweziDol-Oz-sulfide845.407.27709.65122.041,682.1571.06155.356,918.3550.05126.150.6631259
KADI9 165.9CCBClassic Cu-CoKolweziDol-sulfide695.176.50622.7780.491,313.0056.71116.984,947.2339.71108.450.7328161
KADI19 165.9CCBClassic Cu-CoKolweziDol-sulfide1,081.339.54980.05171.591,986.3090.41170.577,638.8252.83101.910.8732671
KADI19 165.9 (NW)CCBClassic Cu-CoKolweziDol-Qz-sulfide590.745.17593.07158.851,155.5945.5222.424,332.1134.2662.040.8428560
DD030 221.56 magCCBClassic Cu-CoKinsanfuMgs3.370.5518.40139.49178.811.100.2320.340.1610.260.6228372
KTO565 553.1CCBClassic Cu-CoKolweziMgs162.072.37130.49185.38422.6811.87111.661,160.897.20128.520.5736478
KOV558 199CCBClassic Cu-CoKolweziDol (in breccia)57.250.482.6112.0657.0820.004.29138.290.3347.540.52936598
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiOz252.782.9022.96ndndnd15.39471.052.8717.810.39370307
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiOz1,436.109.85139.5663.79237.59nd18.402,342.8722.2095.361.23238225
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiCcp30.093.234.543.3528.71ndnd63.250.19149.040.67768564
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiCcp214.541.4617.1523.93557.90nd14.48445.902.66522.571.70378281
K343_214.8DomesVein-hosted Cu (Mo-Au)KansanshiPy207.951.3118.766.17109.49nd4.56431.391.90111.421.07512381
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiCal271.622.4539.1735.41378.57nd24.57581.152.9449.590.65445321
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiCal723.883.98108.83117.13197.53nd5.471,936.9412.7930.020.60341197
PZ-Ba calcitePolandStratiform CuKupferschieferCal734.833.8923.6558.34900.26nd6.412,254.6519.5991.531.26259130
PZ-vein cpyPolandStratiform CuKupferschieferCcp382.052.1213.4524.25481.31nd5.251,287.0813.19213.941.05220101
PZ-vein bornitePolandStratiform CuKupferschieferBn1,203.396.5244.1971.75888.97nd3.803,755.5636.10405.240.91235116
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweCrl463.644.18123.5338.79373.8519.5422.991,653.699.38172.410.86397172
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweCrl613.265.99150.0433.44722.6922.4116.912,209.818.51697.580.91585251
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol35.830.003.6229.5840.960.390.4388.680.1210.810.411,6091,003
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol50.480.153.4932.3633.200.480.50101.140.1912.890.551,221940
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol40.170.001.8559.2545.341.690.2981.68nd18.260.53  
SampleRegionDeposit typeDepositMineralNa (ng)NH4 (ng)K (ng)Mg (ng)Ca (ng)Sr (ng)Ba (ng)Cl (ng)Br (ng)SO4 (ng)Cat/AnCl/BrNa/Br
NS137 cpyZCBClassic arenite-hosted Cu-CoChibuluma WestCcp-Crl-Dol81.105.118.735.4717.46ndnd74.920.81204.000.83209349
NS137 1281 cpy/pyZCBClassic arenite-hosted Cu-CoChibuluma WestPy-Ccp139.775.1448.254.09185.910.622.01286.061.02457.590.98631475
NS137 1300.5ZCBClassic arenite-hosted Cu-CoChibuluma WestCcp-Crl108.841.8210.072.441.62ndnd111.001.1584.191.08217328
NC-car (NW)ZCBClassic arenite-hosted Cu-CoNchangaCrl369.8116.71176.371.7829.410.673.93875.794.73232.140.77417272
KN18 1163.3ZCBClassic argillite-hosted Cu-CoKonkola NorthDol272.431.9571.41106.31195.254.562.81591.993.4982.850.66383272
Mindola dupZCBClassic argillite-hosted Cu-CoMindolaDol-Anh121.731.16104.2783.95373.821.65nd209.561.64536.801.09289259
E1099BX14CCBClassic Cu-CoEtoileCct (in vein)348.012.3117.601.741.85nd1.98500.891.0964.171.031,0331,107
KYA219cpyCCBClassic Cu-CoKamoyaCcp209.216.83198.7440.89473.5119.8433.661,548.9812.94400.110.8027056
H3410 carCCBClassic Cu-CoKisanfuCrl217.365.27374.7633.30195.6813.0439.311,421.3811.220.000.6728667
KOV558/294.6CCBClassic Cu-CoKolweziCrl104.710.6742.1740.08198.614.7520.18673.555.6953.410.7926764
Luilu22 213CCBClassic Cu-CoKolweziCrl-Bn88.301.2053.8373.9570.855.4136.78416.982.3939.931.22394129
KTO559 421.5CCBClassic Cu-CoKolweziCrl-Bn48.790.1733.760.7838.171.770.65248.841.4413.100.68390118
KTO565 543.9CCBClassic Cu-CoKolweziCct136.170.2651.250.15104.816.597.38861.396.731.910.5228870
KTO565 553.9CCBClassic Cu-CoKolweziMixed sulfide98.541.05111.5539.30458.1611.2462.24748.024.27261.650.6239580
KTO565 548CCBClassic Cu-CoKolweziMixed sulfide412.971.9864.3432.19130.143.284.88996.512.9655.600.60758485
KTO565 543.9 dol/sulCCBClassic Cu-CoKolweziDol-Cct-Crl96.080.84125.8584.59420.834.0751.29767.465.7131.821.3330359
KADI19 165.9CCBClassic Cu-CoKolweziDol-Oz-sulfide845.407.27709.65122.041,682.1571.06155.356,918.3550.05126.150.6631259
KADI9 165.9CCBClassic Cu-CoKolweziDol-sulfide695.176.50622.7780.491,313.0056.71116.984,947.2339.71108.450.7328161
KADI19 165.9CCBClassic Cu-CoKolweziDol-sulfide1,081.339.54980.05171.591,986.3090.41170.577,638.8252.83101.910.8732671
KADI19 165.9 (NW)CCBClassic Cu-CoKolweziDol-Qz-sulfide590.745.17593.07158.851,155.5945.5222.424,332.1134.2662.040.8428560
DD030 221.56 magCCBClassic Cu-CoKinsanfuMgs3.370.5518.40139.49178.811.100.2320.340.1610.260.6228372
KTO565 553.1CCBClassic Cu-CoKolweziMgs162.072.37130.49185.38422.6811.87111.661,160.897.20128.520.5736478
KOV558 199CCBClassic Cu-CoKolweziDol (in breccia)57.250.482.6112.0657.0820.004.29138.290.3347.540.52936598
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiOz252.782.9022.96ndndnd15.39471.052.8717.810.39370307
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiOz1,436.109.85139.5663.79237.59nd18.402,342.8722.2095.361.23238225
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiCcp30.093.234.543.3528.71ndnd63.250.19149.040.67768564
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiCcp214.541.4617.1523.93557.90nd14.48445.902.66522.571.70378281
K343_214.8DomesVein-hosted Cu (Mo-Au)KansanshiPy207.951.3118.766.17109.49nd4.56431.391.90111.421.07512381
K112_113.0DomesVein-hosted Cu (Mo-Au)KansanshiCal271.622.4539.1735.41378.57nd24.57581.152.9449.590.65445321
K104_126.4DomesVein-hosted Cu (Mo-Au)KansanshiCal723.883.98108.83117.13197.53nd5.471,936.9412.7930.020.60341197
PZ-Ba calcitePolandStratiform CuKupferschieferCal734.833.8923.6558.34900.26nd6.412,254.6519.5991.531.26259130
PZ-vein cpyPolandStratiform CuKupferschieferCcp382.052.1213.4524.25481.31nd5.251,287.0813.19213.941.05220101
PZ-vein bornitePolandStratiform CuKupferschieferBn1,203.396.5244.1971.75888.97nd3.803,755.5636.10405.240.91235116
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweCrl463.644.18123.5338.79373.8519.5422.991,653.699.38172.410.86397172
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweCrl613.265.99150.0433.44722.6922.4116.912,209.818.51697.580.91585251
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol35.830.003.6229.5840.960.390.4388.680.1210.810.411,6091,003
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol50.480.153.4932.3633.200.480.50101.140.1912.890.551,221940
Shink 83-308CCBRoan Gp-hosted U-Ni-CoShinkolobweDol40.170.001.8559.2545.341.690.2981.68nd18.260.53  

Cl/Br and Na/Br are molar ratios; Cat/An is the ratio of total cation equivalents to anion equivalents

Abbreviations: Anh = anhydrite, Bn = bornite, Cal = calcite, CCB = Congolese Copperbelt, Ccp = chalcopyrite, Cct = chalcocite, Crl = carrollite, Dol = dolomite, Mgs = magnesite, nd = not determined, Py = pyrite, Qz = quartz, ZCB = Zambian Copperbelt

Contents

References

REFERENCES

Annels
,
A.E.
,
1979
,
Mufulira greywackes and their associated sulphides
:
Institution of Mining and Metallurgy Transactions
 , v.
88
, p.
B15
23
.
Annels
,
A.E.
,
1989
,
Ore genesis in the Zambian Copperbelt, with particular reference to the northern sector of the Chambishi basin
:
Geological Association of Canada, Special Paper 36
 , p.
427
452
.
Armstrong
,
R.A.
,
Robb
,
L.J.
,
Master
,
S.
,
Kruger
,
F.J.
, and
Mumba
,
P.A.C.C.
,
1999
,
New U-Pb age constraints on the Katangan sequence, Central African Copperbelt [abs.]
:
Journal of African Earth Sciences, Special Issue
 , v.
28
, p.
6
7
.
Ball
,
S.M.
,
2015
,
Hydrothermal nickel sulfide hosted in Neoproterozoic carbonate and evaporitic rocks of the Menda Central prospect, Democratic Republic of Congo
: Unpublished M.Sc. thesis,
Golden, Colorado
,
Colorado School of Mines
,
124
p.
Barra
,
F.
,
Broughton
,
D.
,
Ruiz
,
J.
, and
Hitzman
,
M.W.
,
2004
,
Multi-stage mineralization in the Zambian Copperbelt based on Re-Os isotope constraints[abs.]
:
Geological Society of America Abstracts with Programs
 , v.
36
, p.
516
.
Barron
,
J.W.
,
2003
,
The stratigraphy, metamorphism, and tectonic history of the Solwezi area, Northwest Province, Zambia
:
Integrating geological field observations and airborne geophysics in the interpretation of regional geology
 : Unpublished Ph.D. thesis,
Golden, Colorado
,
Colorado School of Mines
,
233
p.
Bartholomé
,
P.
,
Evrard
,
P.
,
Katekesha
,
F.
,
Lopez-Ruiz
,
J.
, and
Ngongo
,
M.
,
1973
,
Diagenetic ore-forming processes at Kamoto, Katanga, Republic of the Congo
, in
Amstutz
,
G.C.
, and
Bernard
,
A.J.
, eds.,
Ore in sediments
 :
New York
,
Springer-Verlag
, p.
21
41
.
Batumike
,
M.J.
,
Cailteux
,
J.L.H.
, and
Kampunzu
,
A.B.
,
2007
,
Lithostratigraphy, basin development, base metal deposits, and regional correlations of the Neoproterozoic Nguba and Kundelungu rock successions
,
Central African Copperbelt: Gondwana Research
 , v.
11
, p.
432
447
.
Bernau
,
R.
,
Roberts
,
S.
,
Richards
,
M.
,
Nisbet
,
B.
,
Boyce
,
A.
, and
Nowecki
,
J.
,
2013
,
The geology and geochemistry of the Lumwana Cu (± Co ± U) deposits, NW Zambia
:
Mineralium Deposita
 , v.
48
, p.
137
153
.
Brems
,
D.
,
Muchez
,
Ph.
,
Sikazwe
,
O.
, and
Mukumba
,
W.
,
2009
,
Metallogenesis of the Nkana copper-cobalt South orebody, Zambia
:
Journal of African Earth Sciences
 , v.
55
, p.
185
196
.
Broughton
,
D.W.
, and
Rogers
,
T.
,
2010
,
Discovery of the Kamoa copper deposit, Central Africa Copperbelt, D.R.C.
:
Society of Economic Geologists Special Publication 15
 , v.
1
, p.
287
298
.
Broughton
,
D.W.
,
Hitzman
,
M.W.
, and
Stephens
,
A.J.
,
2002
,
Exploration history and geology of the Kansanshi Cu-(Au) deposit, Zambia
:
Society of Economic Geologists Special Publication
 
9
, p.
141
153
.
Bull
,
S.
,
Selley
,
D.
,
Broughton
,
D.
,
Hitzman
,
M.
,
Cailteux
,
J.
,
Large
,
R.
, and
McGoldrick
,
P.
,
2011
,
Sequence and carbon isotopic stratigraphy of the Neoproterozoic Roan Group strata of the Zambian copperbelt
:
Precambrian Research
 , v.
190
, p.
70
89
.
Byrne
,
G.M.
,
2017
,
Geometries and genesis of breccias within the halokinetic Kaulukundi diapir at the Mashitu mine, Katanga, Democratic Republic of Congo
: Unpublished MSc. thesis,
Golden, Colorado
,
Colorado School of Mines
,
146
p.
Cahen
,
L.
,
François
,
A.
, and
Ledent
,
D.
,
1971
,
Sur l’age des uraninites de Kambove Ouest et de Kamoto Principal et révision des connaissances relatives aux minéralisations uranifères du Katanga et du Copperbelt de Zambia
:
Annales de la Société Géologique de Belgique
 , v.
94
, p.
185
198
.
Cailteux
,
J.
,
1977
,
Particularites stratigraphiques et petrographiques du faiceau inferieur du Groupe des Mines au centre de l’arc cuperifere Shabien
:
Annales de la Société Géologique de Belgique
 , v.
100
, p.
55
71
.
Cailteux
,
J.
,
1986
,
Diagenetic sulphide mineralization within the stratiform coppercobalt deposit of West Kambove (Shaba-Zaire). Sequence of mineralization in sediment-hosted copper deposits
(part
2
), in
Friedrich
,
G.H.
, et al
, eds.,
Geology and metallogeny of copper deposits
 :
Berlin, Heidelberg
,
Springer-Verlag
, p.
398
411
.
Cailteux
,
J.
,
1991
,
La tectonique intra-Katanguienne dans la region nord-ouest de l’arc Lufilien (Shaba, Rép. Du Zaire)
:
Annales de la Société Géologique de Belgique
 , v.
113
, p.
199
215
.
Cailteux
,
J.
, and
Lefèbvre
,
J.J.
,
1975
,
Stratigraphie et minéralisations du gisement cuprifère de Kipapila, Shaba, Zaire
:
Annales de la Société Géologique de Belgique
 , v.
98
, p.
317
330
.
Cailteux
,
J.
,
Binda
,
P.
,
Katekesha
,
W.M.
,
Kampunzu
,
A.B.
,
Intiomale
,
M.M.
,
Kapenda
,
D.
,
Kaunda
,
C.
,
Ngongo
,
K.
,
Tshiauka
,
T.
, and
Wendorff
,
M.
,
1994
,
Lithostratigraphic correlation of the Neoproterozoic Roan Supergroup from Shaba (Zaïre) and Zambia, in the central African copper-cobalt metallogenic province
:
Journal of African Earth Sciences
 , v.
19
, p.
265
278
.
Cailteux
,
J.L.H.
,
Kampunzu
,
A.B.
,
Lerouge
,
C.
,
Kaputo
,
A.K.
, and
Milesi
,
J.P.
,
2005
,
Genesis of sediment-hosted stratiform copper-cobalt deposits, Central African Copperbelt
:
Journal of African Earth Sciences
 , v.
42
, p.
134
158
.
Capistrant
,
P.L.
,
Hitzman
,
M.W.
,
Wood
,
D.
,
Kelley
,
N.M.
,
Williams
,
G.
,
Zimba
,
M.
,
Kuiper
,
Y.
,
Jack
,
D.
, and
Stein
,
H.
,
2015
,
Geology of the Enterprise hydrothermal nickel deposit, North-Western Province, Zambia
:
Economic Geology
 , v.
110
, p.
9
38
.
Cathles
L.M.
,
Oszczepalski
,
S.
, and
Jowett
,
E.C.
,
1993
,
Mass balance evaluation of the late diagenetic hypothesis for Kupferschiefer Cu mineralization in the Lubin basin of southwestern Poland
:
Economic Geology
 , v.
88
, p.
948
956
.
Chabu
,
M.
, and
Boulègue
,
J.
,
1992
,
Barian feldspar and muscovite from the Kipushi Zn-Pb-Cu deposit, Shaba, Zaïre
:
Canadian Mineralogist
 , v.
30
, p.
1143
1152
.
Chi
,
G.
, and
Savard
,
M.M.
,
1997
,
Sources of basinal and Mississippi Valleytype mineralizing brines: Mixing of evaporated seawater and halite dissolution brine
:
Chemical Geology
 , v.
143
, p.
121
125
.
Cosi
,
M.
,
De Bonis
,
A.
,
Gosso
,
G.
,
Hunziker
,
J.
,
Martinotti
,
G.
,
Moratto
,
S.
,
Robert
,
J.P.
, and
Ruhlman
,
F.
,
1992
,
Late Proterozoic thrust tectonics, high-pressure metamorphism and uranium mineralization in the Domes area, Lufilian arc, northwestern Zambia
:
Precambrian Research
 , v.
58
, p.
215
240
.
Darnley
,
A.G.
,
1960
,
Petrology of some Rhodesian Copperbelt orebodies and associated rocks
:
Institute of Mining and Metallurgy Transactions
 , v.
69
, p.
137
173
.
Darnley
,
A.G.
,
Horne
,
J.E.T.
,
Smith
,
G.H.
,
Chandler
,
T.R.D.
,
Dance
,
D.F.
, and
Preece
,
E.R.
,
1961
,
Ages of some uranium and thorium minerals from east and central Africa
:
Mineralogical Magazine
 , v.
32
, p.
716
724
.
Decrée
,
S.
,
Deloule
,
É.
,
De Putter
,
T.
,
Dewaele
,
S.
,
Mees
,
F.
,
Yans
,
J.
,
Marignac
,
C.
,
2011
,
SIMS U-Pb dating of uranium mineralization in the Katanga Copperbelt
:
Constraints for the geodynamic context: Ore Geology Reviews
 , v.
40
, p.
81
89
.
de Magnée
,
I.
, and
François
,
A.P.
,
1988
,
The origin of the Kipushi (Cu, Zn, Pb) deposit in direct relation with a Proterozoic salt diapir; Copperbelt of Central Africa, Shaba, Republic of Zaire
:
Society for Geology Applied to Mineral Deposits, Special Publication
 
5
, p.
74
93
.
Demesmaeker
,
G.
,
François
,
A.
, and
Oosterbosch
,
R.
,
1963
,
La tectonique des gisements cuprifeères stratiformes du Katanga
, in
Lombard
,
J.
, and
Nicolini
,
P.
, eds.,
Gisements stratiformes de cuivre en Afrique, 2ème partie: Tectonique: Association des Services Géologiques Africains (Paris), Lusaka
 
1962
, p.
47
115
.
Derriks
,
J.J.
, and
Vaes
,
J.F.
,
1956
,
The Shinkolobwe uranium deposits: Ccurrent status of our geological and metallogenic knowledge
:
International Conference on the Peaceful Uses of Atomic Energy
,
United Nations
,
Geneva
, August 8–20, 1955, Proceedings, v.
6
, p.
94
128
.
Dewaele
,
S.
,
Muchez
,
Ph.
,
Vets
,
J.
,
Fernandez-Alonzo
,
M.
,
Tack
,
L.
,
2006
,
Multiphase origin of the Cu-Co ore deposits in the western part of the Lufilian fold-and-thrust belt, Katanga (Democratic Republic of Congo)
:
Journal of African Earth Sciences
 , v.
46
, p.
455
469
.
Duffett
,
M.
,
Sebagenzi
,
S.
, and
Selley
,
D.
,
2010
,
Gravity and magnetics of the Central African Copperbelt
:
AMIRA P872—Sediment-hosted copper deposits of Congolese, Zambian and central Australian basin systems, Final Report
 , p.
49
78
.
Eglinger
,
A.
,
Ferraina
,
C.
,
Tarantola
,
A.
,
André-Mayer
,
A.S.
,
Vanderhaeghe
,
O.
,
Boiron
,
M.C.
,
Dubessy
,
J.
,
Richard
,
A.
, and
Brouand
,
M.
,
2014
,
Hypersaline fluids generated by high-grade metamorphism of evaporites
:
Fluid inclusion study of uranium occurrences in the western Zambian Copperbelt: Contributions to Mineralogy and Petrology
 , v.
167
, p.
967
994
.
Eglinger
,
A.
,
Vanderhaeghe
,
O.
,
André-Mayer
,
A.S.
,
Goncalves
,
P.
,
Zeh
,
A.
,
Durand
,
C.
, and
Deloule
,
E.
,
2016
,
Tectono-metamorphic evolution of the internal zone of the Pan-African Lufilian orogenic belt (Zambia)
:
Implications for crustal reworking and synorogenic uranium mineralizations: Lithos
 , v.
240
, p.
167
188
.
El Desouky
,
H.A.
,
Muchez
,
P.
, and
Cailteux
,
J.
,
2009
,
Two Cu-Co phases and contrasting fluid systems in the Katanga Copperbelt
,
Democratic Republic of Congo: Ore Geology Reviews
 , v.
36
, p.
315
332
.
Fleischer
,
V.D.
,
Garlick
,
W.G.
, and
Haldane
,
R.
,
1976
,
Geology of the Zambian Copperbelt
, in
Wolf
,
K.H.
, ed.,
Handbook of stratabound and stratiform ore deposits
 : v.
6
, p.
223
352
.
François
,
A.P.
,
1974
,
Stratigraphie, teconique et minéralisations dans l’Arc cuprifére du Shaba (Rép. du Zaïre)
, in
Bartholome
,
P.
ed.,
Gisements striformes et provinces cupriféres: Centernaire de la Société Géologique de Belgique, Liège
 , p.
79
101
.
François
,
A.P.
,
1986
,
La parties centrale de l’Arc cuprifere du Katanga
:
étude géologique: Tervuren African Geoscience Collection
 , v.
109
,
61
p.
François
,
A.P.
,
1994
,
La structure tectonique du Katanguien dans la region de Kolwezi (Shaba, Rep. du Zaire)
:
Annales de la Société géologique de Belgique
 , v.
116
, p.
87
104
.
François
,
A.P.
, and
Cailteux
,
J.
,
1981
,
La Couverture Katangienne entre les Socles de Zilo et de la Kambompo, République du Zaire—région de Kolwezi
:
Tervuren, Belgique, Musée Royal de l’Afrique Centrale, Annales Serie In 8o, Sciences Géologiques, no
 .
87
,
50
p.
Fuchs
,
L.
,
Schmeling
,
H.
, and
Koyi
,
H.
,
2011
,
Numerical models of salt diapir formation by down-building: The role of sedimentation rate, viscosity contrast, initial amplitude, and wavelength
:
Geophysical Journal International
 , v.
186
, p.
390
400
.
Halley
,
S.W.
,
Wood
,
D.
,
Stoltze
,
A.
,
Godfroid
,
J.
,
Goswell
,
H.
, and
Jack
,
D.
,
2016
,
Using multielement geochemistry to map multiple components of a mineral system
:
Case study from a sediment-hosted Cu-Ni camp, NW Province, Zambia: Society of Economic Geologists Newsletter
 , no.
104
, p.
1
,
15
21
.
Halpin
,
J.
, and
Selley
,
D.
,
2010
,
Insights to geometry and structural evolution of the Central African Copperbelt from zircon provenance analysis
:
AMIRA P872—sediment-hosted copper deposits of Congolese, Zambian, and central Australian basin systems, Final Report
 , p.
33
47
.
Hammerli
,
J.
,
Rusk
,
B.
,
Spandler
,
C.
,
Emsbo
,
P.
, and
Oliver
,
N.H.S.
,
2013
,
In situ quantification of Br and Cl in minerals and fluid inclusions by LA-ICP-MS
:
A powerful tool to identify fluid sources: Chemical Geology
 , v.
337–338
, p.
75
87
.
Hanson
,
R.E.
,
Wardlaw
,
M.S.
,
Wilson
,
T.J.
, and
Mwale
,
G.
,
1993
,
U-Pb zircon ages from the Hook granite massif and Mwembeshi dislocation: Constraints on Pan-African deformation, plutonism, and transcurrent shearing in central Zambia
:
Precambrian Research
 , v.
63
, p.
189
209
.