“Hydromet technology has advanced significantly over the past 15 years and is being applied to areas where it was never thought to be viable,” said Don Stevens, general manager, operations, Long Harbour Processing Plant.
More and more processors are moving toward full hydrometallurgy or incorporating hydromet steps in conventional pyrometallurgical facilities, as is the case in Sudbury, he noted.
Conventional pyrometallurgy requires two separate facilities – a smelter and a refinery, whereas “we do it all in one step,” said Stevens. “We dissolve the feed directly and refine and recover it without having to handle and transport it to a different facility, so it’s more efficient.”
Hydrometallurgy is also more energy efficient, noted Stevens.
“We did an assessment in 2002 to propose hydrometallurgy as an environmentally responsible new process and did a lifecycle analysis of total energy required for smelting and refining, as well as for hydrometallurgy, and we found that our hydrometallurgy process used significantly less energy – mostly because of the one plant, versus the two plants and the efficiencies gained there.”
Another important benefit of hydrometallurgy is the elimination of air emissions.
“That’s huge,” said Stevens. “That’s the big advantage that we saw in hydromet in the first instance. In a smelter, you’re effectively burning all the sulphur that’s associated with the feed and then recovering it at great expense. With hydrometallurgy, the sulphur still needs to be removed from the nickel by reaction with oxygen, but we do it under water in a large pressure vessel called an autoclave. Instead of going up a stack and scrubbed, it remains behind in a solid phase.”
Vale’s initial thinking was to build a conventional smelter and refinery in Newfoundland. At the time, the company was working on the assumption of a mine and mill producing about 20,000 tonnes a day, and a mine life of 10 years based on the then known reserves and resources.
“However, a panel looking at the mine and mill project from a environmental assessment perspective didn’t want to see a project that had a very short duration with all the problems associated with a boom-bust scenario, so they essentially directed that the mine and mill rate be scaled down to 6,000 tonnes per day, which would extend the operating life of the mine and the known reserves at that point in time to 17 or 18 years.”
At 6,000 tonnes per day, conventional smelter and refinery technology was deemed uneconomic, forcing Vale to go back to the drawing board.
The company had already committed to building a smelter and refinery in the province as part of a development agreement with the government of Newfoundland and Labrador, so as an alternative to going the pyrometallurgical route, it promised to invest in an R&D program to develop a hydromet process. In the event that the R&D program wasn’t successful, it committed to building a conventional refinery in Newfoundland to process nickel matte produced from Voisey’s Bay concentrate at smelting complexes in Sudbury or Thompson, Manitoba.
“We went back to our labs in Sheridan Park, Mississauga to look at some known extraction methods and put together a flow sheet that we felt would be less expensive both in terms of an initial investment and also long-term operating costs so we could justify going forward and developing the property,” said Stevens.
Vale invested $200 million in an R&D program, going from bench testing to a 1:10,000 scale mini pilot plant at Sheridan Park and, ultimately, to a 1:100 scale demonstration plant in Argentia, Newfoundland. The demonstration plant cost $100 million and operated for two and a half years at an additional cost of $50 million.
“It’s a very complicated flow sheet,” said Stevens. “Our feed is not just nickel. It’s nickel, cobalt and copper, plus all the various impurities. We had to make sure we had an effective and proven method to not only extract and recover the metals, but also to track and manage the deportment of all the impurities and trace elements.
“In a smelter, most of the things you don’t want end up in a slag phase and it’s easily skimmed off away from the nickel melt, but because we’re dissolving everything, or most things, we have to design or find best practices to recover the trace elements to make sure they don’t build up.
“It’s one of the key reasons we built the demonstration plant in Argentia. We learned how these minor elements build up over time and we invented or deployed new processing steps to specifically deal with them.”
Small quantities of silica, magnesium and manganese would “poison the process unless we found ways to extract them,” said Stevens.
The technology was pronounced technically and economically feasible by the end of 2008 and Vale began construction of the full-scale 50,000 tonne per year hydromet plant in 2009 at Long Harbour, 40 kilometres north of the demonstration plant.
Hydromet technologies have long been used to process nickel laterite ore and copper. Canadian miner Sherritt International Corp. operates a hydromet plant in Fort Saskatchewan, Alberta that processes nickel and cobalt sulphide ore from its joint venture Moa operation in Cuba, but the Sherritt process wasn’t suitable for Newfoundland, said Stevens.
“What we’re doing isn’t done anywhere else,” he remarked.
Vale has been shipping Voisey’s Bay concentrate to Sudbury and Thompson for smelting and refining since the mine went into production in 2005, but under its agreement with the government of Newfoundland and Labrador, it will have to find 440,000 tonnes of concentrate from outside the province to make up for it.
Late last year, the company announced it would close its Thompson smelter by 2015 because of a shortage of feed linked to the startup of the Long Harbour facility. At its Sudbury smelting and refining complex, which will also cease processing Voisey’s Bay concentrate in 2013, Vale is proposing to spend $2.2 billion on an Atmospheric Emission Reduction capital program to meet tougher government emission standards (see story in June 2011 issue of Sudbury Mining Solutions Journal).
10 million person hours
The Long Harbour project represents a huge economic boost for Newfoundland and Labrador. A total of 1,260 workers were onsite in early July and the number was due to peak at 2,200 workers through the summer. In total, the project will generate some 10 million person hours of employment during the construction phase and employ 475 people when in operation.
The structures are now erected, cladding is in progress and in May, Vale awarded a two-year $600 million contract to KBAC Constructors, a Kiewit/BMA/G.J. Cahill Partnership, to supply mechanical, piping, electrical and instrumentation services, the largest contract Vale has awarded in Canada.Engineering, procurement and construction management services are being performed by Texas-based Fluor Corp.