The global demand for metallic mineral resources such as Ni, Cu, and platinum-group elements (PGEs), despite reductions related to events such as recessions, is projected to continue its overall rise. This is mainly fueled by the industrial growth in countries such as China and India but will continue to be buoyed by not only Asian countries but also African and South American nations, as well as continued demand in North America and Europe. Economic development coincides with an increased demand for resources of all types, and world demand will grow at a time when sustainable development is a key issue that must be thoughtfully considered if we are to preserve Earth for future generations.

In line with these challenges, mineral resource companies must expand our proven mineral resources and maintain an acceptable profit margin for new production facilities. Environmental protection will continue to be a priority in the development of new properties for both production and beneficiation. To meet demand and comply with environmental practices that assure responsible development, exploration and mining companies must continue with programs to enhance strategies for the discovery and production of new metallic resources that can be developed with minimal environmental disturbance and are in line with well-conceived plans for remediation.

In this paper I review two tenants of exploration strategies for Ni-rich, Cu- and PGE-bearing sulfide deposits that represent significant additions to past guidelines. The changes encompass a widened view of geologically prospective target environments and a changing strategy toward mining and remediation costs. I will also review new extraction technologies that promise to bring large-tonnage but low-grade resources in a variety of environments into the range of economic viability.

It remains clear that high-grade, large-tonnage deposits, such as those of the world famous Sudbury and Noril’sk Nidistricts (Figs. 1, 2) will continue to be exploration prizes. However, the probability of locating high-tonnage and high-grade deposits appears to be relatively low based on the number of significant discoveries in the past 30 years. The Voisey’s Bay deposit is generally regarded as the most significant greenfields nickel discovery since the Australian Yilgarn craton discoveries nearly 30 years ago (e.g., Kerr, 2003). World-class deposits are commonly defined as those in the top 10 percent of contained metal for a particular commodity. This definition is, at times, a problematic one because some very low grade deposits are high tonnage (e.g., Mount Keith, Australia, Figs. 1, 2) with large contained metal content. Greenfields efforts to find such deposits are perhaps the number one priority of all exploration companies seeking Ni and Cu resources. In general it is believed that large volume magmatic systems (e.g., flood basalt provinces) are advantageous for the production of high-tonnage ore deposits because of the potential for immiscible sulfide liquid to extract metals from a large volume of magma. Within a favorable magmatic system deposits of smaller size, but high-grade, have also proven to be of significance based primarily on the proximity of several individual low-tonnage deposits. Proterozoic komatiite-related deposits, such as those in the Raglan area of Quebec (Cape Smith belt; e.g. Lesher, 2007), are excellent examples of where the close proximity of several shallow and low-tonnage but high-grade sulfide lenses have lead to an overall favorable situation for economic development (Figs. 1−3). Individual deposits along the ˜60- to 70-km strike length of the Raglan Formation typically range from 2 to 8 million metric tons (Mt), grading in excess of 2 percent Ni, but the estimated cumulative resource along the belt is now estimated to exceed 60 Mt (Goldbrook Ventures, 2010).

A deposit such as that at Voisey’s Bay, Newfoundland and Labrador, provides an example of a small footprint magmatic system that hosts a moderate- to potentially large-tonnage resource. The Voisey’s Bay deposit (Fig. 4) is a conduit-style deposit located in an olivine gabbro-troctolite intrusion (e.g., Naldrett and Lightfoot, 1999; Naldrett et al., 2009), part of the Nain plutonic suite that includes many large anorthositic bodies (Ryan, 2000). The Ovoid deposit, one of at least three mineralized areas that constitute the Voisey’s Bay deposit, contains 32 Mt of proven and probable reserves at 2.8 percent Ni, 1.8 percent Cu, and 0.14 percent Co and is now being mined by an open-pit operation (e.g., Huminicki and Sylvester, 2007). Total reserves and resources are estimated to exceed 141 Mt at ˜1.6 percent Ni (Government of Newfoundland and Labrador, 2010) and include disseminated sulfides in troctolitic rocks located on either side of the Ovoid deposit (Eastern Deeps and Reid Brook, Fig. 5). The disseminated deposits of the Eastern Deeps may one day be mined by underground methods, but the smaller Ovoid deposit (Fig. 5) has underpinned the development of the mine through a comprehensive impact benefits agreement with the native populations and the province of Newfoundland and Labrador. Because of weather in the North Atlantic, concentrate must be shipped for treatment using an ice-resistant ship, but costs are high. Transportation costs are vital concerns in the development of small-tonnage orebodies and the potential for exploitation of much larger tonnage disseminated ores at Voisey’s Bay was a positive factor. Some of the largest costs of the operation at Voisey’s Bay are related to the need to fly workers to a remote location and the need to power the operations with diesel (P. Lightfoot, writ. commun., 2010).

Geologically the Voisey’s Bay deposit represents a magma conduit and/or chamber sequence generated by the open-system passage of troctolite-generating tholeiitic magma (e.g., Li et al., 2000; Naldrett et al., 2009). Dense sulfide liquid concentrated in widened parts of the conduit, including near the entrance to large chambers such as where massive sulfide is present below the Eastern Deeps (Fig. 6). The conduit environment offers all of the requirements that are normally thought to be necessary for the development of economic orebodies. These include the involvement of large volumes of mafic and/or ultramafic magma, the potential for the interaction of magma with sulfur-bearing country rocks, and favorable localities in the magmatic plumbing system, in terms of fluid dynamic considerations, where sulfides may collect, such as where the “feeder” dike widens as it opens into the Eastern Deeps chamber (e.g., Naldrett, 1999; Ripley and Li, 2003). The accumulation of sulfide in the dike connecting the larger chambers at Voisey’s Bay illustrates the strong potential for Ni sulfide ore genesis in this type of environment. The Proterozoic Tasiuyak Gneiss country rocks (Fig. 5) are locally sulfidic and graphitic and are thought to have supplied S to passing magmas (e.g., Ripley et al., 1999, 2002; Ryan, 2000; Naldrett et al., 2009). Interaction of magma with S-bearing country rocks at Voisey’s Bay occurred when Ni was available to enter a sulfide melt and was not exhausted by incorporation in olivine. Irvine (1975), Lightfoot and Hawkesworth (1997), and Li and Ripley (2005, 2009) have also shown that contamination of a tholeiitic magma by siliceous country rocks in the absence of additional sulfur can lead to sulfide saturation. Holwell et al. (2007) and Lehmann et al. (2007) have suggested that CO₂ derived from carbonates can also promote the attainment of sulfide saturation in a mafic magma. In summary, the conduit environment is where magma interaction with country rocks must occur, whether that occurs in deeper staging chambers or along immediate conduit walls.

The huge Noril’sk deposits in Siberia (Figs. 1, 7) were perhaps generated in magma conduit systems related to Permo-Triassic rifting and the formation of large volumes of flood basalts (e.g., Naldrett et al., 2009, and references therein). Controversy exists as to whether or not the ore-bearing sills actually supplied magma to overlying flood basalts (e.g., Latipov, 2002; Arndt et al., 2003), but there is no doubt that the subvolcanic intrusions are part of an extensive magmatic system where the potential interaction with country rocks was maximized (e.g., Grinenko, 1985; Naldrett et al., 1996; Lightfoot and Hawkesworth, 1997; Keays and Lightfoot, 2007; Li et al., 2009). Obviously the Noril’sk district illustrates that very high tonnage and high grade sulfide ores may be generated in a subvolcanic conduit environment and this potential has driven millions of dollars worth of exploration in other rift-related flood basalt provinces. It is unclear if the Voisey’s Bay system was part of a feeder system for overlying volcanic rocks that may have been eroded away (e.g., Ryan, 2000), but the deposits at Voisey’s Bay are related to a conduit that links two different magma chambers. The PGE enrichment at Noril’sk may be indicative of the passage of large volumes of magma through a conduit (e.g., Brugmann et al., 1993; Li et al., 2009), but Ni- and Cu-rich sulfides may form without the involvement of such large volumes of magma.

The Eagle deposit in northern Michigan represents a potential extreme in terms of the feasibility of mining low-tonnage but high-grade deposits. Geologists from Kennecott Minerals (now Rio Tinto) utilized models of Noril’sk and Voisey’s Bay in their exploration protocol, which led to the drilling of olivine-rich targets in the Marquette-Baraga dike swarm to the east of the axis of the ˜1.1 Ga Midcontinent Rift System (Figs. 1, 8, 9). Extensive drilling outlined a small, high-grade deposit in widened parts of one of the dikes or intrusions (now known as the Eagle intrusion, Fig. 10). Although disseminated sulfide mineralization is present, only massive and semimassive (net-textured) mineralization has been considered in reserve plus resource calculations. Currently the deposit is listed at 4.05 Mt with an average grade of 3.57 percent Ni, 2.91 percent Cu, 0.73 g/t Pt, and 0.47 g/t Pd (Michigan Government, 2010). Mine permits are in the final stages of approval and extraction of ore is to commence in 2011. Excellent transportation routes and the proximity of the Sudbury smelter complex have contributed to the decision to mine the deposit. In addition, underground mining of the localized sulfide accumulation ensures a small environmental footprint; this was a particular concern because of the sandy, glaciated overburden that supports a diverse flora and the transport of meteoric water to the aquifers in the area. Taken together, the characteristics of the Eagle deposit and nearby infrastructure have proven favorable for economic development of the deposit as a source of additional concentrate for a nearby smelter.

The conduit environment of the Eagle intrusion has been the subject of extensive investigation by Ding et al. (e.g., 2010, in press). Eagle occurs in the Midcontinent Rift system that is associated with large volumes of mafic flood basalts. Mineralogic and geochemical studies indicate that contamination by country rocks was important for the attainment of sulfide saturation and that multiple sources of country-rock sulfur were involved in the genesis of immiscible sulfide liquids. Parental magmas were similar in composition to those of picritic lavas that are present in the Lake Superior area (Fig. 11) within the earliest volcanic rock sequences associated with initiation of rifting (˜1108 ± 1 Ma; Fig. 9). Those of Group 1 of the Mamainse Point Formation (e.g., Berg and Klewin, 1988) are particularly similar, in terms of both major and trace elements. Olivine and pyroxene, as well as sulfide, were concentrated in the widened areas of the conduit (Fig. 10). Isotopic studies indicate that reaction between passing magmas and accumulated sulfides and silicates has promoted exchange reactions and controlled isotopic signatures that reflect the exchange with less contaminated magma. This process is expected where passage of early-stage magmas armors conduit walls and prevents further contamination by country rocks (e.g., Ripley and Li, 2003).

The conduit-type deposits represent a style of mineralization that may generate very large deposits, such as those at Noril’sk, or small, high-grade occurrences, such as those at Eagle. The latter types will be exploited if necessary transportation routes and infrastructure are either in place or can be developed relatively inexpensively. Because the orebodies can be mined by underground methods, the environmental impact of the mining process may be reduced, particularly if ground-water interaction with the deposit is minor. The flood basalt province of the Midcontinent Rift system is an environment where conduit-style deposits of the Noril’sk-type should be explored; what the success at Eagle demonstrates is that small intrusions or dikes may be excellent targets and should not be overlooked. It is feasible that several small-tonnage deposits may exist along the length of a dike swarm, for example, and collectively these could contribute to a large-tonnage deposit, not unlike the Proterozoic komatiites of the Raglan area.

In recent years, most exploration for magmatic Ni-Cu-PGE deposits was focused on extensional environments and plume-associated rifting or on volcanic belts where komatiites were either known to occur or were suspected. The reason for these choices was the premise that large-scale crust-mantle processes, associated with high degrees or large volumes of mantle melting, were required to generate world-class deposits. The convergent margin environment, where many types of hydrothermal deposits are located along present or past active continental margins, was not considered to be attractive for the generation of magmatic Ni-Cu-PGE deposits. The relatively lower volumes of ultramafic rock types that are found in subduction environments, as well as the low abundance of Ni-rich olivines in many arc-related basalts, supported this assessment. Fluids derived from subducting slabs that are responsible for mantle wedge metasomatism may mobilize Ni, and hence sub-arc mantle is typically proposed to be less Ni rich than rift-related mantle. More recent work has questioned many of these assumptions, and deposits such as those at Aquablanca, Spain (Tornos et al., 2001; Pina et al., 2006) and within the Tati and Selebi-Phikwe belts of Botswana (Maier et al., 2008) have been interpreted to indicate that convergent tectonic settings may be favorable for Ni-Cu-PGE deposit genesis. When arc magmatism is viewed in concert with partial melting of the mantle wedge, there is little reason to exclude this environment for Ni-Cu-PGE exploration. Magmas generated in the mantle wedge may be hydrous and contaminated with components inherited from the downgoing slab, but Ni-bearing magma should be produced at moderate degrees of mantle melting. Crustal conduit systems and staging chambers must exist where interaction with country rocks occurs. The processing of potentially large volumes of magma through the conduit system to overlying volcanic arcs can generate high-grade Ni-Cu-PGE deposits. In addition to Aquablanca and the deposits from Botswana mentioned above, the Kalatongke deposit in China (287 ± 5 Ma, Han et al., 2007) has been interpreted as a conduit-style occurrence emplaced in Lower Carboniferous tuffs and shales of the Cordilleran-type Central Asian orogenic belt (Song and Li, 2009). The metal-hosting mafic intrusions show strong evidence for derivation from metasomatized subarc mantle. The small size and conduit-like features of the intrusions are similar to those of the Eagle deposit, but the environment of emplacement differs considerably.

Mafic and/or ultramafic intrusive rocks located above subduction zones may be found in what are known as Ural-Alaskan complexes. Igneous lithologic units that include dunite, olivine clinopyroxenite, and hornblende-bearing ultramafic rocks were described from the southeastern Alaskan panhandle by Taylor and Noble (1969) and Taylor (1967; Figs. 1, 12). Because of their similarity to a group of intrusions that occur along the Ural Mountains of Russia, they were referred to as Ural-Alaskan intrusions. These intrusions are perhaps the best-known examples of ultramafic rocks that occur within subduction zone environments (e.g., Irvine, 1974; Johan, 2002). They are frequently zoned with cores of dunite rimmed by the pyroxene- and hornblende-rich rock types. Concentric zoning is not well developed in some intrusions (e.g., Duke Island, Irvine, 1974; Thakurta at al., 2008a) and origins as layered complexes related to mafic magma fractionation have been proposed as opposed to origins involving melt-rock interaction (e.g., Keleman and Ghiorso, 1986) and emplacement as parts of mantle diapirs (e.g., Burg et al., 2009). If the complexes involve relatively high degrees of mantle melting to produce magmas that generate the olivine-rich rock types, then the potential for the development of Ni-rich sulfides should also exist. Several Ural-Alaskan intrusions have been identified that host Ni-Cu-PGE sulfide mineralization. The Salt Chuck mine in Alaska produced 0.3 Mt of sulfide ore between 1906 and 1941 (Loney and Himmelberg, 1992; Watkinson et al., 2002;), although Cu, Au, Ag, and Pd were the principal products with little or no production of Ni. Sulfide occurrences are also reported from the Quetico intrusions of the western Superior province of Canada (Pettigrew and Hattori, 2006), the Akarem gabbroic complex in Egypt (Helmy and Mogessie, 2001), the Turnagain Complex in British Columbia (Nixon, 1998), and the Duke Island Complex in Alaska (Thakurta et al., 2008a).

Ural-Alaskan intrusive complexes, at Turnagain and Duke Island (Figs. 1, 12), contain zones of sulfide mineralization and may illustrate the factors necessary for the generation of Ni-Cu-PGE sulfide mineralization. Ural-Alaskan intrusions, in general, have been linked to Pt-rich placers that are spatially related to the intrusions (e.g. Urals Platinum belt, Russia; Goodnews Bay, Alaska). Although the exact source of Pt in the placers is generally not known, it is most commonly thought to reside in PGE-bearing chromitites associated with the dunites of the Ural-Alaskan complexes (Johan, 2002). Sulfide minerals are, in contrast, notably sparse in these intrusions and in downstream alluvial deposits. One hypothesis of why little or no sulfide mineralization is found in the intrusions is that the oxidized nature of many arc magmas would not favor sulfide as a predominant sulfur species; instead, sulfur would be present as sulfate (e.g., Thakurta et al., 2008b). Studies of sulfide zones in the Duke Island (Fig. 12) and Turnagain complexes by Nixon (1998), Thakurta et al. (2008a), and Scheel et al. (2009) have shown assimilation of both country-rock sulfur and carbon. The carbon has lowered the fO₂ of the magma and stabilized sulfide. At Duke Island, the contamination occurred after extensive olivine crystallization at a time when clinopyroxene was becoming a liquidus phase. This led to the development of relatively Ni poor but Cu and PGE rich sulfide melt. The Ni-depleted olivine in some dunites suggests that Ni-rich sulfide liquid may have formed in localized staging chambers, below the presently exposed intrusions. In the case of Turnagain, net-textured, Ni-rich sulfides occur in dunites. Hard Creek Nickel has plans to develop a resource of 580,000 t of recoverable Ni that grades only 0.2 percent Ni (Hard Creek Nickel, 2010). This is possible due to advances in extraction technology, as described below. Geochemical and isotopic data (Nixon, 1998) from the Turnagain deposit are similar to those collected at Duke Island and indicate that carbon assimilation has been important for reducing fO₂ of the source magma and generating conditions favorable for the collection of sulfide liquid. Whether assimilation of country rocks has occurred at other sulfide-bearing Ural-Alaskan intrusions is unclear, but this process takes place in crustal environments above subduction zones. Hence, the tectonic environment should not be overlooked in future Ni-Cu-(PGE) exploration programs because mafic to ultramafic complexes may occur in geologically complex convergent margin settings. Geophysical and geochemical techniques (e.g., O’Reilly et al., 2009) have proven successful in identifying ancient convergent margins and should be useful in the search for Ural-Alaskan and similar intrusions that could host significant sulfide-rich Ni-Cu-(PGE) mineralization.

Large-tonnage, low-grade Ni-Cu-(PGE) deposits in mafic and/or ultramafic rocks are in production at Mount Keith and Noril’sk but in many instances have proven economically marginal when evaluated in a traditional flotation concentrate-smelter-refinery sequence. Such an example would be the deposits in troctolitic and/or gabbroic rocks of the ˜1.1 Ga Duluth Complex, which are located near the basal contacts of the South Kawishiwi and Partridge River intrusions with underlying Archean to Proterozoic metasedimentary, metavolcanic, and plutonic rocks (Figs. 1, 9, 13). The deposits are characterized by disseminated sulfides grading ˜0.6 percent Cu and <0.2 percent Ni. Plans for combined open-pit and underground mining showed only marginal economic viability. This was true although individual deposits were large; due to the large combined size of the deposits, the Duluth Complex ranks as one of the largest repositories of magmatic Cu, Ni, and PGE in the world (Fig. 2; Naldrett, 2009). In terms of contained copper, the Duluth Complex rivals Noril’sk, as well as giant porphyry Cu deposits, such as those at Bingham and Butte. Only the Bushveld Complex, the Great Dyke, and Noril’sk contain more tonnage of PGEs than does the Duluth Complex. Individual deposits along the ˜40-km strike length of the mineralized basal contact contain from ˜50 Mt to 1 Bt of indicated resources (Fig. 13).

New hydrometallurgical extraction techniques involving high-pressure chloride and sulfuric acid treatments will be, in large part, likely responsible for bringing the Duluth Complex deposits into production (e.g., Duluth Metals, 2010; PolyMet Mining, 2010). Unlike the Eagle deposit, for example, where Ni grades are high enough to easily contribute feed to existing smelters at Sudbury, the low-grade ores in the Duluth Complex could not have been previously concentrated with assurance of long-term economic viability. The pressure-leach methods that have been applied to nickel laterites have been modified to treat sulfide ores with exceptional success. The PLASTOL™ process of the SGS Group (SGS Group, 2010) produced recoveries in excess of 97.6 percent for Cu, Ni, Pt, and Pd in tests of PolyMet Mining’s NorthMet deposit in the Duluth Complex (Fig. 13). PolyMet has announced plans for a two-stage development of their NorthMet deposit, with construction of new hydrometallurgical and electrowinning plants to be partially funded from initial sales of flotation concentrates (PolyMet Mining, 2009). Hard Creek Nickel’s plan for development of the Turnagain deposit, described above, is also aligned to the pressure-leaching process of Ni recovery. Developments in process technology will continue to be of particular importance in the development of ore systems with greater than ˜500,000 t of contained Ni, such as the Turnagain deposit. If the autoclave system of concentration can replace smelting, then low-grade, large-tonnage deposits, such as those in the Duluth Complex, may have long and profitable production futures.

Greenfields efforts to locate world-class Ni-Cu-PGE deposits are a priority for large Ni-Cu companies and exploration in terranes characterized by large volumes of mafic magmatism will continue. However, over the past 15 years it has become clear that very economic orebodies may be associated with small footprint magmatic systems. Exploration in the Midcontinent Rift system, for example, certainly has the potential to locate orebodies the size of those at Noril’sk. However, smaller conduit-style deposits, such as Rio Tinto’s Eagle deposit that occurs as part of the rift system, represent high-grade targets that are capable of supplying high-quality feed to pre- existing smelters. The development of small, underground mines is advantageous in terms of environmental disruption and remediation. Vale Inco’s Voisey’s Bay deposit in Labrador illustrates that small footprint, conduit-style deposits have the potential to link massive and disseminated sulfide occurrences with a combined resource in the 100- to 200-Mt category (Vale Inco, 2010). Small-tonnage, high-grade deposits may also occur along a structure or lithologic trend, and, when taken together, can constitute a world-class deposit belt. The Proterozoic komatiite-related deposits in the Raglan area are such examples, but conduit-related deposits in dike swarms, such as at the Eagle deposit, also offer potential for similar spatially related deposits.

Requirements for the generation of world-class deposits include the involvement of large quantities of magma, the potential for magma interaction with country rocks, and suitable geometries for the collection of immiscible sulfide liquid. Although these prerequisites are perhaps most readily met in areas of continental extension and flood basalt-related volcanism, they can also be met in the suprasubduction zone compressional environment. Deposits such as Kalatongke in China and Aquablanca in Spain illustrate that convergent zone settings may be favorable for sulfide-rich Ni-Cu-(PGE) mineralization. Ural-Alaskan−type intrusions offer a new type of target for Ni-rich sulfide deposits, particularly if assimilation of country rock-derived S and C occurred prior to extensive olivine crystallization. Hard Creek Nickel’s Turnagain deposit in British Columbia represents a low-grade resource (580,000 t of recoverable Ni at 0.2% Ni: Hard Creek Nickel, 2010) in dunite of a Ural-Alaskan intrusion.

Because of new developments in pressure leach extraction of Cu, Ni, and PGEs, deposits like Turnagain are economically viable. Hydrometallurgical methods of Cu, Ni, and PGE concentration have the potential to render large, low-grade deposits, such as those in the Duluth Complex, economically viable. The large tonnages of contained Ni, Cu, and PGEs in the deposits along the base of the Duluth Complex constitute a world-class resource, but only new advances in process technology have raised the potential that there may be alternative methods to traditional floatation-smelting and established new synergies with hydometallury technologies. New methods of ore treatment will certainly prove to be beneficial for many low-grade Ni-Cu-PGE occurrences in the future.

I thank Jim Mungall for the invitation to write this paper. Conversations with Tony Naldrett, Chusi Li, Peter Lightfoot, Dean Rossell, and Steve Hauck are gratefully acknowledged. Comments from Peter and Dean were particularly helpful in evaluating the outlook of large companies with respect to future exploration trends. I would also like to thank Mike Lesher and Peter Lightfoot for formal reviews of an earlier draft of the manuscript. Rich Goldfarb and Erin Marsh are thanked for handling the editorial responsibilities; their comments improved the presentation of the manuscript.