Coke is a very important raw material used in the production of aluminum. A typical smelter uses on a net basis around 350 kg of calcined coke for each tonne of aluminum produced. The world market for calcined coke, including the small volumes used by other industries, is around 25 million tonnes per year.

To get an historical perspective of the supply demand situation for coke it is useful to look at the correlation between oil demand (a rough proxy to coke supply) and aluminum demand (a good proxy for coke demand). Looking at the graph below, it is clear that the end of the last century has been a turning point.

World oil demand and primary aluminum consumption 1990-2012

Until the turn of the century, primary aluminum demand was growing at roughly the same pace as oil supply and therefore the pressure on coke price was linked to the ups and downs of the aluminum cycle. Price of coke over time was relatively stable compared to LME, which means that coke prices were going slightly down in real terms. In that period, the job of finding coke was left to the purchasing department with the view to minimizing the total cost to the system.

From the early 2000, the pace of aluminum demand has accelerated sharply, to around 6% growth y/y, while oil demand has stayed at the historical level of 1% to 1,5%.  This has put a lot of pressure on coke price which has roughly doubled in terms of its percentage to LME. Even in the current recession, coke price relative to oil price has stayed at a very high level compared to LME.

From the relatively simple analysis above one can clearly see that there is only one option for coke prices in the future and that is up.

But how high can they go?

Looking at the equation from a refinery perspective one might think that higher coke prices should translate into higher coke generation. However if we consider the following points one might understand that the coke price increases of the last 7 years have not done much if anything to increase the coke supply.

The first point is that coke price increases have not kept pace with the increase in oil prices. In the 90’s WTI averaged 20$ per barrel. It is today at more than 5 times this level. During the same period calcined coke prices have roughly doubled. For a refinery who thinks in differential terms between oil prices and product prices, the incentive to produce anode grade green coke has therefore deteriorated compared to 15 years ago.

The second point is that most of the North-American refineries (the main producers of green coke in the world apart from China) do not sell calcined coke but green coke. The job of calcining the green coke to supply the smelters with calcined coke is mostly left to independent calciners. In the last 10 years the margin of the calciners have been heavily correlated with coke prices. This means that calciners have benefited more than the refineries from the rise in coke prices, especially at the peak of the market.

The third point is that coke revenues represent such a small fraction of the overall refinery revenues (around 1 to 2%) that a change, even significant, to today’s price of coke will not change significantly the profitability of a refinery and therefore, the volume or quality of the coke it produces.

It follows from the above that a very significant change in the price of green coke to a refinery will be needed to justify an additional process to remove sulphur, vanadium and nickel from the cokes that are today too contaminated to be used by the industry.

It is difficult to be very specific about a price range for coke that would justify an investment in a new process to clean the green coke in a refinery. Each refinery has a specific configuration and would have its own cost benefit analysis to make to justify such an investment. Also, the long term view of many refineries in Europe and North-America is bleak and one might question in this context a significant investment aimed at improving the quality of what is essentially a byproduct. Presumably it would make more sense for a refinery in a low cost country (such as India or China) to look at such a scenario.

A personal opinion is that we will need calcined coke prices consistently above 1000$ before we can see significant adjustments to the coke supply either in quantity or quality.

In such a context a prudent manager needs to think about its coke strategy going forward. In a short supply scenario, the aim of such a strategy should not be to minimize today’s cost of the purchased coke, but rather make sure that the coke will be available long term to the smelter in the required quality and quantity. In today’s economic conditions this is not an easy decision, but for the ones who can afford it, now is the time to make the long term moves.

[Serge Roy has worked almost 30 years at Alcan and Rio Tinto Alcan and is now consultant in business planning for the aluminum industry, specializing in the areas of carbon, recycling and ingot marketing.  Serge Roy can be reached at roy@aluconsult.eu]