Hard-to-find coking coal’s price rockets as global steel demand orbits

20th June 2008 By: Keith Campbell - Creamer Media Senior Deputy Editor

Things, for steelmakers, go better with coke – the coke concerned being, of course, the product used in blast furnaces and electric arc furnaces to produce iron and steel.

Coke is 90% carbon and is light, but strong – it has to be, to bear the weight to which it is subjected in a blast furnace – porous and, when heated, reactive. At the very start of the iron- and steelmaking process, coke fines – also known as coke breeze – are employed in the sintering of the iron-ore, which converts the irregular lumps (iron-ore fines) of the ore into larger pieces between 10 mm and 25 mm. This greatly increases the efficiency of the smelting process.

In blast furnaces, coke plays more than one role – it is the fuel that provides the heat necessary for the smelting process, and it is the source of the carbon monoxide which reduces the oxygen out of the compounds which bind the iron to the ore, separating it, in liquid form (given the heat), to pool at the base of the furnace, from which it can be tapped.

The blast furnace is ‘charged’ with the sintered iron-ore, coke and limestone – this last combines with the impurities (non-iron constituents) of the ore and separates them as slag; this slag floats on top of the liquid iron and so can be skimmed off. In addition to providing the heat and the reduction agent, the coke provides a permeable medium through which the reduction gas can pass.

In electric arc furnaces, coke is not needed to generate heat, but serves to reduce the molten chemical slag in the base of the furnace, created by the electric arc process. Again, this separates the iron from the slag. The molten iron is then taken to another furnace – usually a blast oxygen furnace – and converted into liquid steel. Coke is not employed in this process – the carbonising agency employed in the production of carbon steel is derived from petroleum.

The iron produced in a blast furnace is known as pig iron, and, on average, it requires 1,5 t of iron-ore and 0,45 t of coke to produce 1 t of pig iron. For completeness, it should be noted that coke is used in, and is equally essential to, the smelting of all metals, playing the same roles in every case.

The main alternative to coke would be charcoal, but it requires the destruction of about 4 000 m2 of woodland to make enough charcoal to produce one ton of iron, so this is a highly environmentally destructive option, although it is still used in some parts of the world. “There are very few alternatives to coke, although South Africa is researching alternatives – with some limited success,” states Professor Rosemary Falcon, who holds the Saneri chair for clean coal technologies at the University of the Witwatersrand.

Coke is produced when coal of the right type is heated up slowly in closed ovens in the absence of air. Unfortunately, very few coals are suitable for the production of coke. They are special, high-quality hard coals, and are known as coking or metallurgical coals. In 2005, the coking coal trade represented only 29% of the global hard coal trade.

Eric Finlayson, head of exploration for Rio Tinto, recently remarked that “coking coal is very hard to find – we are scouring the world to find it”. The biggest, highest-quality, most promising deposit they are currently working on – with a delineation drilling programme intended to deliver an inferred mineral resource by the end of this year – is in Mongolia, some 280 km from the nearest railway. It is, however, close to the Chinese market.


Meanwhile, steel production is booming. According to the International Iron and Steel Institute, total crude steel production (in the 66 countries which report to it) for the first four months of this year was 5,6% higher than for the same period last year.For China alone, steel production jumped 9,1%. Global crude steel production increased by 7,5% in 2007 to reach 1 343,5- million tons – the highest level of crude steel production in history. The year 2007 was also the fifth year in a row in which world crude steel production increased by more than 7%. Although there has been a decrease in the growth rate, demand is still strong. China, unsurprisingly, remains the driving force – its crude steel production increased by 15,7% last year, and, if the Asian giant is excluded, world crude steel production would have increased by only 3,3%.

Given the demand, given the scarcity of coking coal, it is not surprising that the price of the latter is soaring. In April, mining giant BHP Billiton revealed that it expected that coking coal prices would rocket this year by 206% to 240% in comparison with prices for last year. BHP operates collieries in Australia in a joint venture (JV) with Mitsubishi of Japan. Also in April, South Korean steelmaker Posco agreed to a 205% increase in the price of the coking coal it buys from Australia, while Japan’s JFE Steel also accepteda more than 200% increase in the coking coal price – from $98/t, to $300/t. JFE is the world’s third- largest steel producer and Posco is the fourth.

The BHP-Mitsubishi JV is the world’s largest supplier to the seaborne coking coal trade, accounting for 28% of world trade, and is Australia’s largest coal-miner. Australia is the world’s largest producer of coking coal, with production of 83-million tons in 2006, followed by Canada (23-million tons), the US, with 19- million tons, and Poland, at one- million tons, while China, Russia and New Zealand combined provided 8-million tons.

Little wonder that steel companies are striving to secure sources of this key input. Late last month, steel magnate Lakshmi Mittal, CE of ArcelorMittal, bought nearly 10% of Australia’s Macarthur Coal and expressed an interest in buying the entire company (although that might not be possible: China’s Citic has 17,6% of the miner).

Mittal had previously bought three coking coal mines in Russia for $718-million. ArcelorMittal is the world’s biggest steel producer. Meanwhile, late last year, Tata Steel, which ranks sixth in the world, bought a 35% share in a coking coal project in Mozambique, for $88,3-million.

South Africa ranks twenty-first in the world in terms of steel production, making 9,7-million tons in 2006 (up from 9,5-million tons in 2005). This amounts to 0,8% of global crude steel production, but 52% of African production. The basic iron and steel industry contributes some 1,4% of South African gross domestic product and 7,9% of the total value of sales by the national manufacturing sector. By far the biggest local producer is ArcelorMittal South Africa (originally Iscor), which has four steelmaking operations in South Africa – Vereeniging, Vanderbijlpark, Newcastle and Saldanha.

It reported liquid steel production of 7,06-million tons in 2006 – falling to 6,38-million tons last year. Ranked number two is Highveld Steel & Vanadium, whose majority share- holder is Russian steel group Evraz, and which produced 863 142 t of crude steel in 2006. South Africa’s other steelmakers do not smelt iron-ore.

One of ArcelorMittal South Africa’s biggest import costs is the import of coking coal – last year, the company imported 64% of its coking coal requirements. The company produces coke in three coke batteries, as they are called, at Newcastle, Pretoria, and Vanderbijlpark. Of these, the one at Newcastle is new (having been commissioned in November 2006), and it produced 426 000 t of coke last year, playing a major role in driving the company’s total production to a record 840 000 t.

The Pretoria coke battery was relined last year. The company sells coke to South African ferroalloy industries.
“South Africa has very little high- quality coking coal,” points out Falcon. South Africa’s main coking coal producer is Exxaro. “We have one mine producing hard coking coal and a second producing semisoft,” reports Exxaro senior research and development engineer Pierre Jordan.

The hard coking coal is mined at Tshikondeni, and the semi-soft at Grootegeluk, both in Limpopo province. Tshikondeni has an annual production of more than 400 000 t, all for local consumption, while Grootegeluk last year produced some two-million tons of semisoft coking coal (in addition to some 16-million tons of thermal coal).

Of Grootegeluk’s coking coal, 1,3-million tons went to ArcelorMittal South Africa and 0,7-million tons was exported. There is also small-scale production of hard coking coal in KwaZulu-Natal, but these are small deposits and suffer from small seams, which make mining more difficult.

“Semisoft coking coal can’t produce good coke on its own, and must be used in a blend, unlike the hard coals,” explains Jordan. These blends see higher-quality coking coals mixed with lower-quality ones. “The practice here, as in many other places, is to import the high-quality coals and blend them with low-quality local coals,” says Falcon. “Determining the optimum blend for each blast furnace is a very scientific process and is not easy.

The coking coals are blended first, and then turned into coke,” elucidates Jordan. “ArcelorMittal South Africa uses a blend of seven coals; other companies use blends of five coals, while Chinese steelmakers use blends of up to 27 coals. It all depends on the availability of coals and the blend you desire.”

To reduce their demand for coke, South African companies use certain anthracites in sintering plants and in ferrochrome production, blended with coke and (in the case of ferrochrome) char – a less mature bituminous coal – and raw coal.

“These South African anthracites are very special, because they possess a horizontal alignment of their molecules, so they break down easily in the hot liquid slags in electric arc furnaces, releasing free carbon,” points out Falcon. “We are exploring in South Africa for more coking coal,” assures Jordan. “But most of the high-quality coking coal found in South Africa is in very narrow seams, making mining very difficult.”

Others are also exploring for coking coal, in KwaZulu-Natal. For example, JSE-listed Miranda Mineral Holdings reports that its four exploration projects – Uithoek, Burnside, Wasbank (these three are contiguous) and Boschhoek-Boschkloof – together contain 120-million tons of coal, much of which is reportedly coking coal. (Another Miranda project, Sesikhona, should start coal production before the end of this year.) Although the seams are narrow, varying between 0,9 m and 1,5 m, the company is confident that the quality of the coal more than compensates for this.

Then there is Mozambique.

“There is a lot of exploration for coking coal in Mozambique,” highlights Falcon. “Mozambique is the most promising area in the region for coking coal,” agrees Jordan, “but the issue is logistics – trans- port difficulties drive up the costs and can render exploitation uneconomic.”

The big coking coal project in Mozambique is, of course, Companhia Vale do Rio Doce’s (Vale) Moatize project. Moatize is located in Tete province, some 1 700 km north of the capital city, Maputo. The project involves an investment of $1,3-billion by Vale, and the project development plan and the mining contract were both approved by the Mozambican government in June last year.

Moatize will be an openpit operation and will produce an average of 8,5-million tons of metallurgical coal, plus 2,5-million tons of thermal coal, each year. The mining contract is valid for 25 years, but the mine is forecast to have a life of 35 years. The mine is expected to be commissioned in the second half of 2010. Moatize is linked to the port of Beira by a railway that has been concessioned to Indian State-owned railway com- panies, which are currently rehabilitating the line.

It is expected to be reopened next year. However, South African industry is unlikely to benefit from Moatize – it is expected that the metallurgical coal will be sent to Brazil, to feed that country’s large and growing steel industry. (The Brazilian steel industry ranked ninth in the world in 2007, producing 33,8-million tons, a 9,3% increase over 2006.)

An example of other coking coal exploration activities in Mozambique is provided by Australian Stock Exchange-listed company Riversdale Mining (which owns the Zululand anthracite colliery, in KwaZulu-Natal, and also has another anthracite project under development in the sameprovince).

Riversdale holds 16 explor-ation tenements, totalling 203 460 ha, in the lower Zambezi coal basin, and seven exploration tenements covering 86 620 ha in Tete province. Significantly, the latter are contiguous to both Riversdale’s other tenements in the country, and Vale’s Moatize coal concession area. It was in Riversdale’s Benga and Tete exploration tenements that Tata took a 35% share.

These cover 24 960 ha and, in terms of the deal, Tata will get a 40% share of the coking coal produced and an option to secure more, on commercial terms. The inferred resource for the area covered by this JV stands at 1,9-billion tons. The primary resource is hard coking coal, with thermal coal as a secondary product. Tata wishes to use the Mozambican coking coal to supply its Corus subsidiary in the UK and Europe.