CO₂ storage: do impurities matter?

Carbon dioxide (CO₂) streams from all capture processes will contain variable levels of impurities, sometimes referred to as co-components or contaminants. The impurities in CO₂ streams originate from the source of CO₂ and will vary if sourced from a power plant vs a cement kiln for example. Common impurities are nitrogen (N₂), oxygen (O₂) and water (H₂O), but also air pollutants such as sulphur (S) and nitrogen oxides (SO_x and NO_x), particulates, hydrochloric acid (HCl), hydrogen fluoride (HF), mercury (Hg), other metals and trace organic and inorganic contaminants. The removal of certain contaminants from the stream may be required for health, safety and environmental protection reasons, but also to ensure efficient transport and storage.

So is the presence of impurities in the CO₂ stream destined to be stored underground a problem? The possible impacts of impurities are reservoir-specific and depend on the mineralogical composition of the rocks and of course the type of impurity and its concentration. Impacts can vary from slight dissolution creating micro-voids, to mineralisation which fills-up the pore-space. Although the potential mechanisms through which certain impurities could affect storage capacity or integrity are well understood, simulating the exact conditions of a storage complex and the gradual accumulation of impurities in the laboratory pose significant problems.    

Research on the effects of these impurities on the integrity of geological storage formations is limited. Most CO₂-rock interactions have been investigated with laboratory experiments using pure CO₂. Only a small number of laboratories world-wide are able to physically simulate the affects of sulphur oxides and acids at high pressure and temperature over geological time frames, and a lack of data around the kinetics of diffusion and likely reactions increases the uncertainty within computer models. Nevertheless, two broad categories of possible impacts on CO₂ storage processes can be identified.

First, the presence of non-condensable gases such as hydrogen (H₂), argon (Ar), N₂ and O₂ can significantly reduce the density of the CO₂ stream. In particular, captured CO₂ streams from oxyfuel capture technologies can contain relatively large percentages of non-condensable gases, up to approximately 10 per cent depending on the capture process. Specifically for storage, the low density of the injected CO₂ stream would lead to inefficient utilisation of pore space, thereby reducing the amount of CO₂ that can be stored within a given reservoir. Therefore, there are economic reasons to remove, at least partially, certain contaminants in the CO₂ stream. 

Second, the presence of certain impurities such as hydrogen sulphide (H₂S), NO_x and SO_x can reduce the pH of the formation water. This has the potential to consequently affect porosity and permeability of the reservoir and also cap rock. For example, the presence of SO₂ will result in the formation of sulphuric acid that has a marked effect at reducing the pH of the reservoir water. The drop in pH can result in an acidified radial zone around the injection point (up to around 200 m), whereby dissolution of carbonate material (limestones and dolostones), if present, could occur. At the periphery of the acidified zone pH increases and the precipitation of secondary minerals may occur. Depending on the reservoir mineralogy the precipitation of secondary minerals is higher in cases where CO₂ is co-injected with SO₂. It has been modelled that after approximately 100 years, sufficient precipitation of secondary minerals may take place and changes may be induced in porosity and permeability. This process could lead to the long-term stabilisation of the injected CO₂ and actually improve the integrity of the storage complex. No current scientific literature indicates the co-injection of SO₂ or H₂S leading to short-term (<10 years) porosity reduction that could hinder injectivity. 

What should policymakers do? In terms of setting limits on impurities for storage purposes, some literature points towards a maximum permissible amount of non-condensable gases in the CO₂ stream of 4 per cent by volume. This figure is understood to reflect an optimum balance between gas conditioning costs and the costs of compression; however any such limits would need to consider the source of CO₂, reservoir mineralogy and overall site characteristics. Concerning the potential for impurities to induce changes in the porosity and injectivity of a storage site, there are no indications that the amounts expected in captured CO₂ streams will reduce the efficiency or integrity of storage.