2.2 CO2 and Water

This section gives a background to the complex interactions between CO2 and water. It first looks at the behaviour of the gas in pure water, and then considers the impact when replaced by sea water. Next the reverse situation is addressed, whereby the impact of a small amount of pure water in CO2 is described, and the potential for the formation of hydrates. Having introduced hydrates, the impact of impurities within the CO2 on hydrate formation is described. Finally, the differences between pure water in CO2 and seawater-in CO2 are described.

The objective of this section is to provide a basis on which the hazards relevant to offshore pipelines and platforms can be understood. These are described in 3.3.

2.2.1 CO2-in-pure water

CO2 dissolves in pure water to form carbonic acid (H2CO3):

CO2 + H2O → H2CO3 → H+ + HCO3- pKa = 6,35

HCO3- → H+ + CO32- pKa = 10,25

The solubility of CO2 in water is 1,45 g/litre at 25 °C and 1 bar. Carbonic acid is relatively weak, and it is impossible to obtain pure carbonic acid at room temperatures. Whilst carbonic acid is described as weak, it still carries the potential to corrode carbon steel pipes:

2Fe + H2CO3 → Fe2CO3 + H2

This leaves the pipeline designer with the choice to either use a more costly pipe material, lining the inside of the pipe with a corrosion-resistant coating, or reduce the water content of the CO2 to a level where significant corrosion will not take place.

2.2.2 CO2-in-sea water

CO2 is also soluble in seawater; this is important for offshore pipelines, the environmental impacts of which are described in Annex B. About 89 % of the CO2 normally dissolved in seawater takes the form of bicarbonate ions, about 10 % as carbonate ions, and 1 % as dissolved gas.

2.2.3 Water-in-CO2

The solubility of pure water in liquid CO2 is difficult to measure at low temperatures: at 10 °C it is approximately 1 000 ppm, and at -21 °C it is 180 ppm10. There are a number of models that can be used to describe the solubility of water in CO2, and a number of these have been analysed with reference to experimental observations11. A graphical presentation of the conclusions of this for water-in-CO2 is provided in Figure 2.15.

Figure 2.15 Solubility of H2O in pure CO2

(source: Choi & Nesic 2011)

Figure 2.1512 indicates that at pressures above ~ 80 bar range, there is reasonable confidence that whichever model is chosen for use, up to 2,1 vol % of water will dissolve in CO2 at 30 °C (i.e. 21 000 ppmv). Whilst it seems that any concentration less than this would not allow free water to appear in pure CO2 such that corrosion of carbon steel pipes would not take place, however, CO2 would cool rapidly in either the ground or on the sea floor, thereby reducing the solubility of water in pure CO2 to less than the 1 000 ppm described.

Another consideration is that CO2 also forms hydrates with water (CO2.6H2O), and as such, water that might normally be dissolved uniformly throughout a liquid can become concentrated at particular points. Likewise, some impurities in a CO2 mixture further impact the ability to hold water in solution resulting in the potential formation of hydrates.

2.2.4 Hydrate formation: pure CO2 and pure water

The small amount of water that might normally be dissolved uniformly throughout liquid can become concentrated at particular points because the CO2 liquid is not bonded to the water molecules, but supports the lattice structure surrounding it. The molecules can agglomerate to form hydrates in the presence of the CO2 (CO2.6H2O). Physically, hydrates are solid, and have the appearance of ice (see Figure 2.16).

Figure 2.16 CO2 hydrate

CO2 hydrates are Type 1 clathrate, and the molecular structure is shown diagrammatically in Figure 2.17. Clathrates are crystalline water-based solids in which small molecules with large hydrophobic collections are trapped inside 'cages' of hydrogen-bonded water molecules. In other words, CO2 clathrate hydrates are compounds in which the host molecule is water and the guest molecule is liquid CO2. Without the support of the trapped molecules, the lattice structure of the CO2 hydrates would collapse into conventional liquid water and dissolve.

Figure 2.17 Type 1 clathrate structure, CO2 hydrate

The generic diagram presented as Figure 2.18 explains where these are predicted to exist for pure CO2. The black squares show experimental data13, and the lines of the CO2 phase boundaries are calculated according to thermodynamic tables. The dark grey region (V-I-H) represents the conditions at which CO2 hydrate is stable together with gaseous CO2 and water as ice.


CO2 (S): Solid CO2 CO2(L): Liquid CO2

CO2(V): CO2 Vapour H2O(V): Water vapour

H2O(S): Solid water (ice) H2O(L): Liquid water

V-I-H: Vapour-ice-hydrate envelope CO2 (SC): CO2 in the supercritical phase

Figure 2.18 Hydrate diagram for pure CO2

The formation of hydrates in pipelines containing CO2 is an important issue to consider when developing the operational methodology for CCS projects. Modelling and experimental work14 has been carried out to confirm and establish the conditions where hydrates can develop in pure CO2. Two experiments were conducted using pure CO2.

The first test used pure CO2 in a saturated environment to show the effect of the presence of free water on the formation of hydrates across a range of temperatures and pressures. The results of the experiments are shown in Figure 2.19 (a) where the grey dotted lines define the boundaries of the hydrate/ice stability zones and non-hydrate zones for saturated conditions. To show the necessity and impact of drying the CO2, a further experiment demonstrated the benefit of using pure CO2 dried to less than 250 ppmv. Figure 2.19 (b) depicts these results showing the hydrate/ice stability zones superimposed over the grey dotted lines of the boundaries under saturated conditions.

In both figures, the yellow triangles represent the range of operating pressures for an offshore pipeline in the winter (100 bar to 190 bar at 4 °C). Note in Figure 2.19 (a) the operating conditions are inside the hydrate zone while in Figure 2.19 (b) the operating conditions are significantly to the right and above of the hydrate zones.

Figure 2.19 (a) Hydrate formation in pure saturated CO2


- Below the blue line the CO2 exists in vapour form

- Above the blue line the CO2 exists as a liquid

- Zones to the right of the grey dotted lines (shaded light green) are non-hydrate zones

Figure 2.19 (b) Pure CO2 hydrate/ice stability zones for a 250 ppmv system


H: area where hydrates are present

VCO2: rich CO2 vapour phase

LCO2: rich CO2 liquid phase

I: area where ice will form

 : Typical offshore pipeline operating range

Grey dotted line: hydrate/ice stability zones in saturated conditions (i.e. in the presence of free water) Red lines: Pure CO2 hydrate stability lines: hydrates (or ice) will form in any area to the left shaded blue, with ice hydrates shaded the darker blue

The regions to the left of the red lines in (b) represent the hydrate and ice stability zones for the drier 250 ppmv pure CO2. The dotted lines in this graph have been superimposed from figure (a) to show the difference between drying the CO2 and leaving it saturated and were not a part of this second experiment.

As already mentioned, the second experiment took place at a selected moisture level that would allow for a level of dryness. If the experiment were to take place under dryer conditions the boundaries for the hydrate/ice stability zones would move further to the left.

Figure 2.19 (b) also shows that a typical winter operating range for a cold water offshore pipeline of 100 bar to 190 bar at 4 °C, as indicated by the two triangles, is significantly to the right and well above the 250 ppmv hydrate stability zones.

Drying the CO2 is critical for corrosion management; however, as shown in (a), if there is free water in the CO2 or the system is not dried sufficiently, the system conditions are inside the hydrate stability zone as defined by the grey dotted line. The predicted water content required for hydrate formation at 190 bar is higher than 1 000 ppmv but lower for 100 bar and 4 °C.

2.2.5 Hydrate formation: impure CO2 and pure water

Impurities within CO2 will further affect the point at which hydrates form. The above experiment was repeated with an unsaturated impure mixture of CO2 dried to 250 ppmv. Figure 2.20 shows a more complex structure within the hydrate/ice stability zones for a mixture of 95,8 % CO2, 2 % H2, 2 % N2, 0,2 % CO. The presence of hydrogen within a mixture indicates that at least one source of CO2 was from a pre-combustion capture facility.

Figure 2.20 Hydrate stability zones for 250 ppmv system and impure CO2 typical of a pre-combustion source


- Below the blue line the CO2 exists in vapour form

- Above the blue line the CO2 exists as both a liquid and a vapour

- The blue/red line defines the bubble points for the mixture and the upper boundary for the two-phase region. Above this the CO2 exists as a liquid


H: the area where hydrates are present

V: area where CO2 is in the vapour phase

LC: area where CO2 is in the liquid phase

I: area where ice will form

 : Typical offshore pipeline operating range

Grey dotted line: hydrate/ice stability zone in saturated condition (i.e. in the presence of free water) Red lines: Pure CO2 hydrate stability lines: hydrates will form in any area to the left shaded blue, with ice hydrates shaded the darker blue

Comparing Figures 2.19 and 2.20, it may be observed that the effect of impurities is to move the point at which hydrates will form to the right (i.e. at higher temperature) for the impurities concerned. Other impurities can have the reverse effect. As previously seen in Figures 2.7, 2.13 and 2.14, the presence of hydrogen significantly impacts the bubble point and phases within the different zones.

There is both a vapour/hydrate region and a liquid/vapour/hydrate region (two-phase region) as predicted by Figure 2.7. In Figure 2.20 the blue/red liquid/vapour line which represents the bubble points for the mixture is relatively flat across a wide temperature range and close to 80 bar. The increase in pressure, for the bubble points of the mixture, not only increases the possibility of hydrate formations in lower pressure pipelines, but also increases the potential for two-phase flow and destructive cavitation within CO2 pumps. Again Figure 2.7 indicates that higher levels of hydrogen would have an even greater impact on these bubble points and must be considered in the design of CO2 capture systems and CO2 specifications for offshore pipelines.

In Figure 2.20, it is important to note as previously stated that where impurities like hydrogen are present, the bubble point pressure is increased such that if the pressure drops below this higher level, the first substance to vaporise is the water which in turn will immediately combine with the CO2 to form hydrates. Maintaining the pressure avoids this and the presence of two-phase flow. Re-establishing the pressure will dissolve the hydrates formed and re-establish single phase flow.

The potential for hydrate formation in liquid CO2 is of particular importance when considering the corrosion potential of liquid CO2 from CCS applications at low air and water temperatures (3 °C15), such as will be experienced offshore. The variables in hydrate formation are temperature, pressure, gaseous impurities within the CO2, and the water concentration.

In conclusion it can be stated that, of the variables, the operator can exercise little control over offshore temperatures or the pressure necessary to transmit the CO2to the sub-sea storage. However, the upstream producers can be required to deliver the CO2 with a controlled level of key impurities, and water content sufficiently low for hydrates not to form.

2.2.6 CO2 solubility in seawater

The processes by which a rising bubble of CO2 will dissolve in seawater are complex and described in detail in technical publications16. Reference 18 provides a phase diagram for a CO2 – seawater mix, reproduced here as Figure 2.21.

Figure 2.21 Phase diagram of the CO2 system and the CO2-seawater system.

In Figure 2.21, the dashed black curves mark phase boundaries for pure CO2 (see Figure 2.2). The phases in the pure CO2 system are marked in black (solid CO2, LCO2, and GCO2 for solid, liquid and gaseous CO2 respectively). The dashed blue curves (and the blue squares and dotted circles almost overlapping the horizontal axis) show the phase boundaries of a CO2-free 'pure' seawater system. The phases in 'pure' seawater system (ice and water) are marked in blue. The double blue near-vertical lines are as a result of the variability of the melting point from pure water to seawater. The solid red-purple curves mark phase boundaries in the CO2-H2O system, and the phases in the CO2-H2O system are marked in red.

Between the double blue lines and the solid red-purple curve, there are two phases: hydrate and liquid water (H&W). Between the red solid curve and the dashed black curve, there are two phases, liquid water and liquid CO2 (W&LCO2). To the right of the purple curve, the two phases are liquid water and gas CO2 (W&GCO2). The green curve shows a measured temperature-depth profile in the ocean.

From Figure 2.21 it can also be seen that for the temperatures (275 K, 1,85 °C) and ocean depths up to 150 m the potential exists for the CO2 to be in the gaseous or liquid phase and for hydrates to form on the outer surface of the bubble. Slightly different mechanisms take place for CO2 in the gas and liquid states dissolving in seawater, complicated by the presence of a thin hydrate layer.

Reference 17 also explains that a rising bubble of CO2 will dissolve in whole or in part in seawater dependent on:

- the temperature of the seawater;

- the amount of CO2 already dissolved in the seawater (Le Chatalier's principle17)

To a lesser extent:

- the size of the bubble18, and

- the depth (pressure) of the seawater.

Because rising bubbles of CO2 will dissolve readily in seawater, the amount of CO2 reaching the surface will be reduced, and this will lessen slightly the potential impact on the surface.

10 SPE formation evaluation, Kyoo Y Song and Riki Kobayyashi, Society of Petroleum Engineers, 1987.

11 Thermodynamic models for calculating mutual solubilities in H2O-CO2-CH4 mixtures A. Austegard, M. J. Mølnvik, SINTEF Energy Research, Trondheim, Norway and E. Solbraa, G. De Koeijer, Statoil Research & Technology, Trondheim, Norway, The Institution of Chemical Engineers, September 2006.

12 From Determining the corrosive potential of CO2 transport pipeline in high pressure CO2–water environments, Yoon-Seok Choi, Srdjan Nešic, Institute for Corrosion and Multiphase Technology, Department of Chemical and Biomolecular Engineering, Ohio University, Athens, OH47501, USA. International Journal of Greenhouse Gas Control 5 (2011) pp788–797.

13 Clathrate hydrates of natural gases, second edition, E.D. Sloan Jr, 1998

14 SPE 123778, Effect of Common Impurities on the Phase Behaviour of Carbon Dioxide Rich Systems: Minimizing the Risk of Hydrate Formation and Two-Phase Flow, A Chapoy, R Burgass, B Tohidi (Hydrafact Ltd & Centre for Gas Hydrate Research, Institute of Petroleum Engineering, Heriot-Watt University), and J M Austell, C Eickhoff (Progressive Energy Ltd). Society of Petroleum Engineers, 2009

15 Ocean Climate Status Report 1999, Fisheries Research Services Report 06/00.

16 Fate of rising CO2 droplets in seawater, Youxue Zhang, Department of Geological Sciences, The University of Michigan, Ann Arbor, Michigan 48109-1005, USA; and Key Laboratory of Orogenic Belts and Crustal Evolution, MOE, School of Earth and Space sciences, Peking University, Beijing, 100871, China. Environmental Science Technology, 2005 pp 7719 – 7724.

17 The principle that if any change is imposed on a system that is in equilibrium then the system tends to adjust to a new equilibrium counteracting the change

18 Reference 17 states that the dissolution rate (or the boundary layer thickness) does not vary significantly with the radius of a CO2 droplet. For example, for a given T and P (278 K and 670 m depth), when the radius of a CO2 droplet varies by a factor of 10 from 5 to 0,5 mm, the boundary layer thickness decreases from 44,2 to 31,6 µm, and the dissolution rate increases by only 40 % from 1,26 to 1,77 µm/s. Hence, if T and P were kept constant, the dissolution rate may be regarded as roughly constant and the radius of a droplet varies roughly linearly with time.