UK United Kingdom

The trouble with aluminium

The Australian aluminium industry is in the doldrums. A high dollar, low prices and Asian competition are threatening the industry, with older plants in New South Wales and Victoria under threat of closure…

Australian researchers are looking for ways to reduce the energy consumption of aluminium smelting. AAP

The Australian aluminium industry is in the doldrums. A high dollar, low prices and Asian competition are threatening the industry, with older plants in New South Wales and Victoria under threat of closure.

The debate about the future of the industry often concentrates on the high energy usage associated with aluminium production. The joke description of aluminium as “congealed electricity” is never far away.

It is true that a lot of energy is required to make Aluminium. CSIRO calculate that the embodied energy (all the energy used to make the material) for aluminium is 211 GJ per tonne, compared to 22.7 GJ per tonne for steel. This huge difference in overall energy required to produce the metal helps explain the enormous difference in the scales of the two industries: last year, nearly 1500 million tonnes of steel was made in the world compared to 40 million tonnes of aluminium.

Of course, aluminium is over three times lighter than steel, which means that energy savings can be made over the lifetime of the metal’s use if aluminium replaces steel in transport. Also, the overall energy picture looks much healthier if your energy source is hydroelectricity.

Some argue that only countries with hydroelectricity should make aluminium. I would argue that would not be a very healthy economic solution for Australia and would reduce us to being a supplier of raw materials. I also think this analysis underplays the environmental impact of dams and ignores the part aluminium smelters play in providing consistent base loads for power generation. In many countries around the world, aluminium smelters provide the impetus for construction of power stations, as the investors know that the smelter will provide 24 hour base load, 365 days per year.

Why is so much energy required to make aluminium?

In simple terms, the strength of the chemical bond between aluminium and oxygen is significantly stronger than the same bond between iron and oxygen. As a result, much more energy is required to split the bond and form the metal.

Energy-intensive to make, aluminium can save energy when it’s used to make lightweight, fuel-reducing transport. Kevin McDonnell

There is another complication that makes the situation worse: the relative chemical stability of the two oxides (alumina and iron oxide) and carbon monoxide. Above around 700°C, oxygen prefers to be bonded with carbon than with iron. This means that if you put coal or charcoal with iron ore above 700°C, you can make metallic iron. This discovery was one of the greatest breakthroughs in the history of the human race and underpins a large proportion of the Australian economy, as Australia is blessed with plentiful and rich sources of iron oxide and carbon (coal).

Unfortunately, this is not the case for aluminium oxide. The same reaction needs around 2000°C before metal is made. And, in fact, at that temperature a lot of aluminium carbide and vapour - waste products - is also made. As you can imagine, operating a furnace at 2000°C with poor yield is not attractive and this route for making aluminium has never taken off. Alcoa is currently trying to develop this route but it is unclear whether the “carbothermic route” to aluminium will ever be economically feasible.

Aluminium is made by dissolving the oxide into molten salts, at around a more modest 960°C, and applying a current to help break down the oxide. This is the Hall-Heroult process, which dominates world aluminium production at the moment. The process typically loses around 50% of the incoming energy as low grade heat, which is due in part to the fact that the salts required to dissolve the oxide are so corrosive that there no practical way to keep the heat in. There are also great challenges in the anode and cathode technology in the Hall-Heroult process that result in large resistance losses.

In general, high temperature electrolytic routes for making metals are the last resort for a metal producer. They suffer from low productivity and relatively high energy losses compared to standard pyrometallurgical processes. The aluminium industry would love to have the equivalent of the modern ironmaking blast furnace, which can produce around 3 million tonnes of metal per year in a small fraction of the volume, compared to the many acres of Hall-Heroult cells required to produce that much metal.

What can be done to lower the energy consumption of aluminium production?

A combination of breakthroughs in materials science, developing advanced control systems and straight-out innovation could make a big difference. I and other researchers around the world have identified existing research projects, if developed and commercialised, could reduce the energy by up to 25% of current usage.

For example, Swinburne University of Technology, University of Wollongong and CSIRO are currently working on new refractory materials that can withstand the corrosive conditions of the Hall-Heroult process and keep the heat in. Excellent progress is being made, and if successfully commercialised these materials could make a serious dent into the 25% overall reduction goal.

Promoting Australian innovation in low-energy aluminium would be good for Australian jobs. AAP

At the University of New South Wales, new anode materials are being investigated that would allow resistance losses in the circuit to be lowered, as well as advanced control systems that are being trialled in collaboration with industry by Associate Professor Jie Bao. Joint research at Swinburne University of Technology and CSIRO is looking to reduce resistance losses in the cells using new designs and computational modelling of the electrical contacts. Mr David Molenaar at CSIRO has developed a unique laboratory system for testing new anode designs.

Even more radical research at University of Auckland and Swinburne University of Technology is looking to “bulldoze” the current technology and bypass the problems of the carbothermic route by developing new chemistry to crack this old chestnut. Dr Rhamdhani (Swinburne) and his team are investigating thsulphur route, whilst, Professor Mark Taylor’s team (Auckland) are looking seriously at a chloride chemistry route.

All of this research, and other projects around the world, are focused on lowering the energy associated with aluminium production. Naturally enough, I would like to see Australia take the lead on these developments.

The current business environment appears to be very negative. But opportunities for innovation abound in times of crisis. Given Australia’s great natural advantages in terms of raw materials, I see some light at the end of the tunnel.

The time to innovate has arrived.

Join the conversation

12 Comments sorted by

  1. Craig Read

    logged in via Twitter

    Thanks for an interesting article Geoffrey.

    It reminded me (once again) of this awesome TED talk:

    "In many countries around the world, aluminium smelters provide the impetus for construction of power stations, as the investors know that the smelter will provide 24 hour base load, 365 days per year."

    One of the things that has confused me about Aluminium processing in Australia, is where we've situated…

    Read more
    1. Luke Weston

      Physicist / electronic engineer

      In reply to Craig Read

      Craig, you've also got to factor in the transport of alumina and graphite anodes and other raw materials into the smelter and the transport of aluminium ingots out of the smelter.

      I'm not sure, but I suspect that at Portland a good portion of this transport is done via sea freight, with the aluminium ingots that go to export being directly loaded out onto the boat.

      Introducing an extra leg of road transport for the aluminium product, or the material inputs, introduces its own costs and energy use which need to be factored in.

    2. Chris O'Neill

      Retired Way Before 70

      In reply to Craig Read

      "Is there a reason why the smelter is on the West coast and not the East coast (closer to our power generators)?"

      Politics. There was a lot of lobbying at the time to put the smelter at Portland. Westernport is much, much closer to the generators but it didn't have the lobbyists that Portland had.

  2. Roger Dargaville

    Research Fellow, Energy Research Institute at University of Melbourne

    As well as trying to reduce the amount of electrical energy required to make Al, it would also be advantageous to be able to develop a process that could be switched on and off easily. As I understand it, you can't switch off the power to a smelter for more than a couple of hours or the process solidifies and ruins the equipment. So, Al smelling has no ability to do demand side management.

    And I believe Portland was chosen because of the deep water port required by the ships.

  3. Michael Block


    why isn't the 'waste heat' being harvested to drive electricity generating turbines?

    1. Luke Weston

      Physicist / electronic engineer

      In reply to Michael Block

      Probably because of the intrinsically low efficiency with which low-grade waste heat, at a relatively low temperature, can be harvested to drive electricity-generating turbines due to the Second Law of Thermodynamics.

    2. Michael Block


      In reply to Luke Weston

      Surely improving the energy efficiency by even 10% is significant. It's more than hot enough to drive a steam turbine

  4. John Newlands

    tree changer

    I don't see any estimate of the likely energy savings from a national soft drink can deposit scheme. My guess is that food retailers are dreading it. Perhaps eventually we will have to fill our own containers from a fizzy drink dispenser which is not the end of the world.

    It would be crazy if the same alumina/bauxite and thermal coal now used in Australia went overseas to meet low wages, lax pollution controls (eg PFCs) and lack of carbon tax. Obviously we should have our own baseload low carbon…

    Read more
  5. Colin MacGillivray

    Architect, retired, Sarawak

    Good article
    "Some argue that only countries with hydroelectricity should make aluminium."
    As an Australian living in Sarawak where there is heaps of hydro (and incidentally Swinburne has a branch) makes me think that the establishment and continuation of some Australian industries is unwise.
    Aluminium with no cheap power.
    Rice or cotton with no cheap water.
    Cars with no cheap labour or big home market.

  6. Derek Bolton

    Retired s/w engineer

    it's great that a lot of work is beoing done to make the process more efficient. But even at max theoretical efficiency Al would still be energy hungry. For years, the Australian Al industry has relied on very cheap (subsidised) power.
    The real problem is that the product is too cheap. With no power subsidies and an appropriate carbon price everywhere in the world it would be much more expensive. It would be used less and recycled more. An alternative would be a successful process for using wind and solar, whether by adapting to variability of supply or by developing cost-effective baseload renewables: heat storage for solar thermal or batteries for PV/wind.