Removal of sulphur dioxide from flue gas

In the drive to protect the environment, reducing the concentrations of harmful chemicals that are released into the atmosphere has become a priority for industries. One key example is the removal of sulphur dioxide (SO₂) from flue gas, which can be formed as a by-product of processing of raw materials, for example. In order to achieve this, a filtering procedure needs to be implemented. However, most existing methods, such as "gas scrubbing", require high input power and a specifically suited operation site. This can be expensive and time-consuming for companies to implement.

We have worked with W.L. Gore and Associates, famous for their waterproof clothing GORE-Tex, on a more desirable and cost-effective method of removal of SO₂. In their Gore Mercury Control System pictured above, flue gas flows through stacks of parallel open channels and diffuses into their porous surface, which consists of a fibrous matrix with multiple microscopic catalytic pellets, where a chemical reaction takes place and converts SO₂ into liquid sulphuric acid. This method requires very low input power, does not need a special operation site, and also generates sulphuric acid "for free". However, these filters exhibit one main problem: as liquid sulphuric acid accumulates, it gradually blocks the filter and limits its subsequent efficiency.

We have developed a multi-scale mathematical model that couples an understanding of how liquid sulphuric acid evolves on the microscale with a device-scale model to predict the global filter properties such as its efficiency over time. Our modelling framework comprises three core components:

• a model for the microscopic flow of liquid sulphuric acid in the device
• a model that accounts for the tendency of sulphuric acid to absorb water from its surroundings (hygroscopy) , which provides another mechanism for the sulphuric acid liquid-layer growth within the filter sheets
• a macroscale model using the theory of homogenisation, which provides averaged equations for the thickness of the liquid sulphuric acid layer and sulphur dioxide concentration at different locations in the filter, while retaining the effect of the underlying microstructure

Our modelling has enabled us to describe the filter performance in terms of key parameters that characterize the system chemistry (for example, the rate of diffusion of the chemicals versus the rate of the chemical reaction) and geometry (such as the length of the filter and the diameter of the open channels through which the flue gas flows). It allows for a prediction of how varying these parameters affects the operation efficiency of the device. The result of this mathematical modelling will inform the future design of filters manufactured by W.L. Gore and Associates to maximize their lifetime, as well as being general enough to be applied to other gases and impurities.