Metal-refining Slag, used widely as rail-bed ballast,

--- originating at the copper-nickel smelters of Sudbury, Ontario, Canada

[445 kb]

Figure 1. Here is a typical example of the slag found as ballast on railway lines (past and present) throughout southern Ontario. Such material is found at the railway yard in Sudbury and far, far beyond: the rail yard is closer to the source, the smelting complexes of the great nickel-mining district. One side has been sawn to provide polished thin sections for textural and mineralogical study. Characteristics by which this is identified in hand specimen as Sudbury slag include: flow textures; bubbles (<1 mm to 10 mm and more); a maroon, submetallic surface patina; substantial "heft" (dense); magnetic. This particular sample was found on the railway track west of Campbellford, in Northumberland county, in 2016, picked from innumerable other examples. The railway has been disused since 1980, and the track was taken up around that date. It is now part of the Trans-Canada Trail network.

"Rock of the Month #192, posted for June 2017" ---

Slag -

a product of industrial processes involving heat and melting, can result from the production of diverse metal and glass materials, and particularly the separation of metals from their ores. As metallurgy has become more sophisticated, so the monitoring of metals lost to slag has increased, and the losses reduced. Nevertheless, it seems that slag is seldom considered a worthy subject for research, beyond the inspection of the waste stream for lost profits.

Slags came to my attention in various ways: curious pieces lying on the ground, often far from any obvious mine site; large piles beside an old smelter in the Atacama desert; ballast on railway lines; and a subset of samples proffered as possible meteorites, the submetallic sheen mistaken for the fusion crust. The nearby rail lines, including the spur line from Havelock to the nepheline syenite mine at Nephton (Blue Mountain), are bedded in Sudbury slag and other, natural rock aggregates. The slag is tough, dense, and apparently chemically stable on a time scale of decades, or more. Petrographic study of some samples (G.C. Wilson, unpublished data, 1999-2003) revealed that chosen samples were composed of one or more layers of wüstite (FeO)-bearing glassy fayalite slag with spectacular nucleation of acicular fayalite from the top of the underlying layer. A cursory microprobe study revealed Ni,Cu-rich sulphide blebs, consistent with the Sudbury origin. Some samples may contain quartzite inclusions, engulfed by the molten, fluid slag prior to solidification and cooling. The slag is largely composed of the skeletal (quenched, that is, rapidly cooled) iron-rich olivine (fayalite) crystallites in a glassy matrix, with abundant wüstite and variable, usually minor sulphide. This kind of slag can be recognized by a metal detector in all-metal mode and discriminated successfully in iron-rejection mode (Wilson, 2003).

Historical moment

Sudbury began as a railway construction village in 1883 (Roussell et al., 1998). The skyline is, from a distance, dominated by the `superstack', a 381-m high smelter chimney, the tallest smokestack in the world, which commenced operations in 1972. Peters (1890) gives a detailed and sobering historical vignette of the metal-recovery process in the early decades of the Sudbury camp. At the time there were just three mines in operation. The Evans mine, one mile southwest of "the Copper Cliff" (sic), worked a massive pyrrhotite body with very high Ni and high Cu values. A roasting ground was constructed on rocky land, formerly heavily wooded. The ore was dumped in a roast-yard nearly 0.5 miles long by 100 feet wide (roughly 800x30 m). The wood-fired roasting removed the greater part of S in the ore (and gave rise to further, chemically-driven defoliation, and eventually to the concern over "acid rain", a term seldom heard today, but prevalent in Ontario in the 1980s). Peters noted that an 800-ton heap would ideally burn for about 60 days. The S content had to be reduced to about 7-8% to make a rich matte, with 27% Cu and 15% Ni. The separation of slag from metal involved use of a more refined fuel, top-quality Pennsylvania coke, brought from the USA by the CPR: 7-8 tons of ore were smelted by 1 ton of coke. Heap roasting was consigned to history after three decades or so: the environmental damage was left for all to see until the 1960s, since when a remarkable rehabilitation and reforestation program has been sustained - the changes over the last 30 years have been very impressive.

Slag, continued

Slag production and disposal methods have changed through the 20th century (Bouillon et al., 1999, pp.25-27). The slag (Roussell et al., 1998, p.221) is used as fill in construction, and is broken and processed into aggregates locally referred to as `Dry Pac' and `50 mm minus'. Slag has long been used as fill beneath structures, especially in the smelter town of Copper Cliff (Roussell et al., 1998).

Some relevant details of metal and slag production will be found in the voluminous literature on the Sudbury mines (e.g., McKague et al., 1980; Matousek, 1989). In a study of Cu and Ni sulphide droplets in fayalite slags, slag samples from Copper Cliff had Cu and Ni values of 0.7% and 0.1%, respectively (Itoh et al., 1995). The nickel (and, in some deposits, copper) contents drive the refining process on the large scale, but these deposits also contain valuable platinum group elements (see Cole and Ferron, 2002). Cobalt is another metal recovered from Cu-Ni-PGE ores: in the early 20th cenury only a few mining centres were equipped to recover it (Drury, 1919).

Over many years, the mining and metallurgical sector has made efforts to reduce losses (to tailings and slag) and to minimize emissions of harmful elements. Sulphur, such a common element in many metallic ores, is chemically reactive and - emanating from sulphide minerals such as pyrite and pyrrhotite - is also the key component in acid mine drainage. Improvements at Sudbury led to the production of tails with <0.4% S (Scales, 1994). Such small blebs of sulphide as occur in slag appear to be armoured by the glassy matrix, so their modest S content will be withheld from the environment until the glass devitrifies and the slag - after some unknown length of time - crumbles.


Bouillon,D, Heale,E, Yearwood,P and Hall,GJH (1999) Environmental geology and land reclamation history of Sudbury. GAC/MAC Field Trip Guidebook A6, Sudbury, 34pp.

Cole,S and Ferron,CJ (2002) A review of the beneficiation and extractive metallurgy of the platinum-group elements, highlighting recent process innovations. In `The Geology, Geochemistry, Mineralogy and Mineral Beneficiation of Platinum-Group Elements' (Cabri,LJ editor), CIM Spec.Vol. 54, 852pp., 811-844.

Drury,CW (1919) Cobalt, its Occurrence, Metallurgy, Uses and Alloys. OBM Ann.Rep. 27 part III, section I, 133pp.

Itoh,S, Choo,RTC and Toguri,JM (1995) Electrocapillary motion of copper and nickel matte droplets on fayalite-based slag surfaces. Can.Metall.Q. 34, 319-330.

Matousek,JW (1989) Nickel smelting at Copper Cliff: the second fifty years. In `All that Glitters: Readings in Historical Metallurgy' (Wayman,ML editor), Metallurgical Society of the CIMM, 197pp., 182-184.

McKague,AL, Norman,GE and Jackson,JF (1980) Falconbridge Nickel Mines Limited's new smelting process. CIM Bull. 73 no.818, 132-141, June.

Peters,ED (1890) The Sudbury ore-deposits. Trans.AIME 18, 278-289.

Roussell,DH, Jansons,KJ and Richards,PA (1998) Urban geology of the city of Sudbury. In `Urban Geology of Canadian Cities' (Karrow,PF and White,OL editors), GAC Spec.Pap. 42, 500pp., 207-224.

Scales,M (1994) How we improve. Can.Min.J. 115 no.1, 33-37, February.

Wilson,GC (2003) A test of metal detectors' ability to resolve classes of `hot rocks', including meteorites. Presentation to the Meteorites and Impacts Advisory Committee to the Canadian Space Agency (MIAC), Regina.

Graham Wilson, 18 March 2017 (draft: and microscopy to follow)

Visit the Turnstone "Rock of the Month" Archives!

or the "Rock of the Month Index",