PolyMet clears hurdles at NorthMet

SITE VISIT

Hoyt Lakes, Minn. — This is the Iron Range, the heart of what was once the biggest steel industry in the world. But if that age is gone, northern Minnesota can hope another kind of mining might come forward to fill the void.

The other kind of mining is what is planned by Vancouver-basedPolyMet Mining(POM-T, PLM-X), a junior company that did not let scale get in the way of an ambitious plan to put the NorthMet copper deposit into production, and to succeed in the Duluth Complex copper district where bigger companies did not.

The NorthMet deposit — once called the Dunka Road deposit — was found in the late 1960s by U.S. Steel, which had taken part in an early land rush in the area. It took a bulk sample in 1971, but the project ran into the Duluth Complex’s mine-killer, difficult metallurgy. Historically, it has been easy enough to make a sulphide concentrate from Duluth material using conventional flotation, but smelters did not find the mixed copper-nickel concentrate desirable. Separating the copper and nickel sulphides was possible, but the expense pushed operating costs over what the project could bear.

But while U.S. Steel lost corporate interest in non-ferrous metals and the project gathered dust, the technological ground was shifting under the project’s economics. In 1998, PolyMet — the reconstituted Fleck Resources, a prominent junior platinum group explorer in the 1980s — reactivated the project, which Fleck had leased from U.S. Steel in 1989. (U.S. Steel had retained a 3% net smelter return, which it sold to a private Texas company as part of a larger transaction in 2004.)

There were fits and starts, and an abortive joint venture with Australian iron ore producer North, but over the last nine years PolyMet has raised the NorthMet project off the economic seabed. One difference to the project has been a plan to use hydrometallurgy — a leaching-precipitation process called Platsol — to bypass the smelting process entirely, making it unnecessary to produce separate copper and nickel concentrates. Another difference was PolyMet’s successful negotiation to buy the mothballed Erie taconite plant, a vast iron ore milling and concentration complex near the deposit, from steelmakerCleveland-Cliffs(CLF-N).PolyMet got the mill for US$3.4 million and 6.2 million shares — less than its scrap value — and was also able to get surface facilities and a short rail line from Cleveland-Cliffs for US$15 million and another 2 million shares. Cliffs now holds about 7% of PolyMet.

The tailings pond at the Cliffs-Erie taconite plant. At proposed rates, the pond can accommodate 28 years of tailings from NorthMet.

BY JAMES WHYTE

The project, which could see production by early next year, jumped two barriers: the difficult metallurgy had a solution, and the enormous capital costs entailed by its sheer scale could be whittled down if it had a large, cheap mill. And mills don’t get much cheaper than US$3.4 million and a block of shares. Mining was never likely to be a problem — the deposit is near surface, and makes an open pit with a low stripping ratio.

The numbers

NorthMet is big — in the Complex, onlyTeck Cominco’s(TCK.B-T, TCK-N)Mesaba deposit has a larger published resource. The most recent number on its resource is 579 million tonnes grading 0.27% copper, 0.08% nickel, and 0.007% cobalt, plus 0.23 gram palladium, 0.07 gram platinum, and 0.03 gram gold per tonne. There is an additional inferred resource of 228 million tonnes averaging 0.28% copper, 0.08% nickel and 0.006% cobalt, with 0.27 gram palladium, 0.08 gram platinum and 0.04 gram gold per tonne.

Based on figures from a final feasibility study in September 2006, that resource has an economic reserve of 249 million tonnes, proven and probable, with grades of 0.28% copper, 0.08% nickel, 0.008% cobalt and 0.3 gram total precious metals per tonne (implying precious metal grades similar to those in the resource). The pit’sstripping ratio was just below 1.5.

That feasibility study put capital costs at US$380 million, about half of which would go into the hydrometallurgical plant. Modifications to the Erie mill would take up another US$62 million, but owning the mill saves at least US$200 million. Mining costs for production at 73,500 tonnes of rock and 29,000 tonnes of ore per day were estimated at US$3.45 per tonne, and processing 29,000 tonnes per day at US$7.70 per tonne. Basing revenue on 3-year average prices between 2003 and 2006, the project had a net present value of US$595 million, discounted at 7.5%, and an internal rate of return just shy of 27%, using a reserve figure that was about two-thirds the size of the current one.

The deposit is in an east-facing gabbro of the Duluth Complex, which intrudes over the older Virginia sediments and Biwabik iron formation. The fairly simple conformation of the rock bodies had historically led geologists to believe the Duluth was a single gabbro “lopolith”–an intrusive body roughly concordant with the surrounding strata, and dipping inward toward its centre. Mapping since the 1960s has shown that was an oversimplification, and the Complex is in fact made up of multiple gabbro intrusions, plus granophyre (a felsic differentiate from a gabbro magma) and anorthosite.

The Virginia, sandwiched between the gabbro and the iron formation, is a metamorphosed sequence of sulphidic sediments, often thought to have been a sulphur source for the Duluth gabbros. While the gabbro carried plenty of metal, the thinking goes, the metal would only have crystallized as sulphides if the gabbro had been able to incorporate sulphurrich country rock, bringing the sulphur content of the melt up to a concentration that would allow pyrrhotite and chalcopyrite to form. The definitive studies also show some sulphur contribution from the melt itself.

In core, the mineralization is typical of the Duluth Complex copper and nickel sulphides — disseminated pyrrhotite and chalcopyrite in a coarse plagioclase gabbro. And, as usual, that simple picture varies a little in detail, but doesn’t really break down.

“We have secondary movement of some of the sulphides into highgrade veins,” says Richard Patelke, PolyMet’s project geologist. “But they don’t drive the mine plan.”

Being quite uniform, the mineralization has been easy to build up at what would be, in other deposits, a fairly sparse drill hole density. The indicated resource holds together at drill spacings around 60 metres. At that spacing, “we know what we’re going to get,” Patelke says.

So with a consistent deposit, and a simple open pit to mine it — one that will mostly start mining in ore — the issues around NorthMet shift to the metallurgy and the permitting.

Metallurgy

PolyMet management think the choice of the Platsol hydrometallurgical process is a safe one, in spite of the bad reputation pressureleach technologies got in the late 1990s, when several Australian laterite projects failed on poor production results from the technology. NorthMet is not, after all, a nickel laterite, and PolyMet does not plan to cut corners. Failed hydrometallurgical projects “went cheap, and with hydromet you don’t gocheap,”saysPolyMet’s president, William Murray. “Every piece of equipment that should be used is being used — we will not compromise on that.”

Set against the conspicuous failures in the Australian laterite fields, demonstration-scale projects on copper-nickel sulphides are working in both Newfoundland (Argentia) and Botswana (Tati Nickel), and several copper mines now use a pressure oxidation process on copper concentrates.

The first part of the NorthMet processing stream is conventional, and has had good results at pilot scale: there are four stages of crushing to take the ore down to 12.5-mm size, followed by rod and ball milling to bring it down to 0.1 mm. The Erie plant would be working at only a third of its capacity to take daily production of 29,000 tonnes per day, and the ore’s work index is about 70% that of
the taconite material the mill had to handle in its iron ore days. Joseph Scipioni, Poly-Met’s chief operating officer and an old Iron Range hand, says of the production rate: “I’ve never run a third of a plant in my life.”

From there, the plan is to float a bulk concentrate that will feed the hydrometallurgical plant. That stage is the project’s processing bottleneck, limited to 29,000 tonnes per day under present plans. Separating the copper and nickel into two concentrates is technically possible at this stage, and hasn’t been ruled out; test work on NorthMet material is under way to see if that process can be optimized. If it can, the material should be marketable to smelters, as it carries virtually no deleterious elements.

Producing concentrates for sale to smelters while the hydrometallurgical plant is being built and commissioned would also let PolyMet have some cash flow at the early stages of production. Tests produced a copper concentrate grading 19.4% and a copper-nickel concentrate with 7.5% copper and 4.4% nickel.

The company expects to get a netback of 60% to 70% of gross metal value for copper and nickel in concentrates, and about 50% of the precious metal value. In selling concentrates, “we’re going to give up quite a lot on precious metals,” says Douglas Newby, PolyMet’s chief financial officer. “It’s not a long-term solution.”

So for that long-term solution, it is the Platsol process that PolyMet has assumed to be the way forward. Platsol works in an autoclave vessel at 225C and 3.4 MPa pressure (about 33 times atmospheric pressure), oxidizing sulphides to soluble sulphates and complexing the precious metals as chlorides. The solution is partly recycled through the autoclave to increase extraction rates, then ferric iron is reduced to ferrous iron by adding sulphur dioxide and the palladium, platinum and gold precipitated.

Extraction rates to the solution in pilot-scale tests were above 98% for both nickel and copper, and in the 90s for palladium, platinum, and gold.

The acidic solution, now containing copper and nickel sulphates, is neutralized with limestone, which reacts to produce synthetic gypsum (which PolyMet hopes to sell to wallboard manufacturers). From there, the copper can be recovered by solvent extraction and electrowinning, a well-established technology in the copper industry. Nickel and cobalt can then be precipitated as hydroxides, which also capture any zinc in the solution. The hydroxide recovered in tests so far has about 30% to 40% nickel, around 2% cobalt, and up to 5% zinc, with trace amounts of iron and copper. There are about six possibleend users for the hydroxide, two of which don’t want zinc in their process.

The other issue, permitting, is an elephant in the room for any project in the United States. Minnesota did ban copper and zinc mining in 1979, and has sometimes been seen as an unfriendly state for mining. That reputation is poorly justified by history, the copper- zinc moratorium notwithstanding, and two things work in PolyMet’s favour — it is in large measure redeveloping a “brown-fields” site, and public attitudes in the Iron Range are strong on mining and on jobs.

“If anything, they’re tired of taking so long to do it,” Scipioni says.

Tailings ponds at the Erie plant have 28 years of capacity at the planned production rate, and are permitted. They are, however, out of compliance owing to seepage, and collecting the water is part of the development plan.

As it is elsewhere in the Great Lakes states, acid-rock drainage is ” the issue for permitting,” says Jim Scott, the project’s environmental manager. The ore will average 1% sulphur; waste rock, while it averages 0.08%, can have up to 6% sulphur content. About 2% of the rock to be moved is “definitely acid-generating,” Scott says; 10% to 15% is potentially acid-generating and the rest is not. PolyMet will go into the permitting process with two years of waste characterization under its belt.

The hydrometallurgical plant is designed to return all water to the process stream, limiting the effluent released to the environment. Tailings wastes from the iron ore plant are inert enough that the pond is a habitat for trumpeter swans, and a contractor was dismissed for fishing for pike and pickerel off a reclamation barge.

The project will also have a 2-year baseline set of surface and ground water quality measurements.

The permitting process will be led by the state, with the federal government largely deferring to state decisions on permitting. Most permits cover water use and discharge, and effects on wetlands, and are largely state responsibilities.

There is a public consultation process, but Scott expects most objections can be answered: “We went into this with our eyes wide open that we were going to be challenged on it.”