To main content

Feeding a gas turbine with aluminum plant exhaust for increased CO2 concentration in capture plant


Aluminium smelting plants typically emit 1.5 tonne of CO2 per tonne of aluminium produced. However, the CO2 concentration in the exhaust is very low, typically below 1%, since large quantities of air are employed for cooling of the electrolytic cells and for cooling of the raw gas before gas treatment. Feeding exhaust gas with this low CO2 concentration to a post-combustion capture plant would be very energy-demanding. Simulations with the SINTEF/NTNU in-house simulator CO2SIM predict an energy penalty of 6.1 MJ/kg CO2 captured, for a CO2 concentration of 0.95%, when using MEA in the capture process.

An alternative solution, which is described in the present paper, is to use the Aluminium plant exhaust as the working fluid in a gas turbine that operates in a natural gas fired combined cycle (NGCC), as illustrated in Figure 1. The aluminium plant exhaust gas is rich in oxygen, and should from that point of view be possible to use for combustion purposes in a gas turbine combustion chamber. The resulting CO2 concentration in the exhaust gas would then be roughly 5%. A calculation with CO2SIM shows that this corresponds to a capture penalty of 3.5 MJ/kgCO2 captured, i.e. an energy penalty reduction of 43%.

A first evaluation was made of what the impact could be for the gas turbine if fed with an aluminium plant exhaust gas saturated with sea water instead of with air. It was concluded that the key issue to investigate further is the combination of the remaining minor fractions of impurities from the aluminium plant in combination with the 100% relative humidity in the gas and the presence of NaCl. The main issues related to using this unusual gas turbine working fluid appear to be hot corrosion and filter saturation. There may also be a risk of corrosion in the gas turbine cooling channels.

An experimental setup was designed to filter a slip-stream of aluminium plant exhaust. A HEPA E12 filter was purchased and mounted in the experimental equipment. The test facility was mounted on site at the Hydro Aluminium plant at Sunndalsøra. Three series of measurements were conducted:
• Background measurements, filtering ambient air
• Filtering exhaust gas, new filter
• Filtering exhaust gas after five weeks of operation
A gas turbine filter is normally operated for three years of operation, the short test period was judged to be reasonable for a very first evaluation of the filter capability and of the experimental facility in itself.

Altogether, the conducted study has shed light on the topics that would require further investigation in order to evaluate whether feeding a gas turbine with aluminium plant exhaust gas is a viable option for reducing the CO2 capture penalty from the aluminium industry to acceptable levels.


Academic lecture


  • Research Council of Norway (RCN) / 193816





  • SINTEF Energy Research / Gassteknologi
  • Lund University
  • SINTEF Industry / Sustainable Energy Technology
  • SINTEF Energy Research / Termisk energi
  • SINTEF Industry / Process Technology
  • Norsk Hydro ASA

Presented at





05.06.2013 - 06.06.2013


Sintef og NTNU



View this publication at Cristin