Dec '09/Jan '10

When muscular metallurgy lost its hold

The early history of copper smelting offers lessons in the need for scientific rigour

By L. Southwick

Basic-lined Anaconda converters at Washoe Reduction Plant, Anaconda, Montana, 1917


The  use of sulphur as a heat source in copper smelting is so ingrained today that the attitudes of early smeltermen seem hard to understand. Back in 1909, sulphur burning had only just begun to be adopted. That year, the pioneering Canadian-born engineer-industrialist James Douglas noted that oxidized ores were considered best since sulphide ores had to be roasted and then smelted using expensive coke for heat.1 Despite being so preferred, many oxide and carbonate ore bodies processed in Arizona suffered significant slag losses. Copper losses amounted to 16 per cent at Copper Queen, 30 per cent at Globe/Clifton and 40 per cent at Morenci.

The chief obstacle to the adoption of sulphur burning was the practice of “muscular metallurgy.” Muscular metallurgy, as its name suggests, was concerned with simple mechanical improvements to extractive metallurgy that often required no more than brute force, as opposed to more advanced chemical improvements that would require a fine understanding of complex chemistry. The “muscular” approach acquired early preponderance due to the unrefined state of metallurgical science, the limited training of metallurgists, and rudimentary laboratory analytical and computational tools. In the early U.S. copper industry, oxide or carbonate copper ores were prevalent and smelting relied on multiple reverberatory and blast furnaces in complicated circuits. Muscular smeltermen believed that the transition to sulphide ores with its attendant costs would be disastrous for the copper industry.

Towards the end of the 19th century, pyrite smelting (using pyrite sulphur as the heat source) provided a partial solution to the problem.2, 3 Later, massive sulphide deposits gained prominence. They necessitated many changes in processing-related chemistry and led to a move away from brute-force methods. Yet, some early sulphide smelters wasted the heat contribution from burning sulphur by roasting first. It was only with further improvements in smelting (especially in converting) that the benefits of sulphide ores began to be fully realized.

Until smelters were capable of making the best use of the sulphur, there was no viable alternative between suspending operations or running them at high losses. “Had not the shareholders been willing to accept hopeful promises in lieu of dividends,” wrote Douglas, “the Southwestern smelters would have gone under.”

Conducting tests for Rio Tinto in 1878, the Englishman John Hollway developed the foundational principals of the pneumatic (Bessemer) converting of sulphide ores.4,5 Guided by a pamphlet Hollway had published, the American smelterman Lawrence Austin, who understood the potential of pyrite smelting, conducted his own tests from 1888 to 1893 for several Montana and Colorado smelters. The work of such early pioneers epitomized the struggle to translate concepts to economically viable practice.

Some Montana- and Colorado-based smelters had been roasting their extremely pyritic ores prior to smelting. Austin, who was one of the organizers of the Tolston works, eliminated the roasting step and smelted using just a blast furnace. Because he lacked consistent ore samples and was hampered by defective apparatus and insufficient capital, Austin could not run his tests smoothly and systematically. Consequently, there were many gaps in his data acquisition.

Patchy as they were, Austin’s results were encouraging. Operations at Montana and Idaho established that his process had merit. However, the Boulder Valley ore was too siliceous to flux properly, causing the furnace to remain mostly idle. Nonetheless, the results were good enough for Austin to patent his process and begin selling it to other smeltermen.

Robert Sticht, who himself made notable contributions to pyrite smelting, followed Austin’s work with a critical eye. In his 1895 review,6 Sticht argued that Austin’s results called for much more rigorous inquiry. Austin, in his haste to bring the process to market, had been negligent on several counts. There were very few analyses; dust losses were not properly quantified, sampled or assayed; and overall production balances and yields were impossible to complete. Sticht felt “some professional hesitancy about giving publicity to [such] results.”

Sticht’s critique displays the methodical analysis that characterized the new generation of metallurgists who were struggling to bring order to the world of muscular metallurgy. In addition to lax reporting, Sticht indicted Austin for using imperfect equipment, performing short-duration tests, treating small amounts of ore and not pushing the furnaces to their full capacity. He also criticized the lack of technical skill and the neglect of chemistry.

Despite the flaws in its early development, pyrite smelting was quick to catch on because its benefits were very obvious. With many adopting and refining it early on, pyrite smelting technology developed and was accepted quite rapidly.

Despite setting Austin and many others in motion, Hollway’s work set was a failure. It did, however, contribute to two major developments — pneumatic converting and pyrite smelting. Douglas referred to Hollway’s tests as the most thoroughly reported failure ever published.7

Books and articles from this period often feature better fundamental descriptions of process than much modern scientific literature. They can greatly assist us in learning about how processes work and help metallurgists troubleshoot operating problems.

Finally, we must bear in mind that Sticht’s criticisms are as relevant today as they were over a century ago. All too often, the wrong equipment is used, data collection and evaluation are abbreviated or incomplete, units are not pushed sufficiently hard, and operations are too short-lived to unearth limits and potential problems.


Larry Southwick is a registered engineer consulting in technology development. This article is based on a presentation delivered at the August 2009 COM conference in Sudbury, Ontario.

Post a comment

Comments

PDF Version