Measurement and Modelling of Gas Permeability and Solubility in Polymers for Offshore Pipelines

Several techniques of Enhanced Oil Recovery (EOR) emerged during the last decades to increase oil extraction levels in existent reservoirs1. Among these techniques, CO2 injection is one of the most promising. CO2 is also produced as a by-product in many industries, especially in the energy sector. This greenhouse gas raises concerns regarding its concentration in the atmosphere.

Thus, it should ideally be stored or recycled. Therefore, it is an intrinsic advantage to use CO2 as an EOR gas, since the storage location of the CO2 can be the reservoir itself. Oil reservoirs are increasing in depth, pressure and temperature, beyond 3000 m, where the pressure can exceed 1000 bar and the temperature can be higher than 423 K.

Since the injected gas needs to match the oil reservoir conditions of pressure and temperature, the fluid must be transported at such extreme conditions. Fluid transportation is an often-overlooked step in the process between capture and storage of CO2. Flexible pipelines represent an economical solution compared with rigid pipelines, because they adapt better to the field layout, they have faster and safer installation and have better chemical and mechanical resistance2.

Because of these advantages, they are a key component in the oil and gas industry, especially for offshore applications. A flexible pipeline consists of different layers of materials, including both polymer and metal. There are two polymeric layers: the first is located at the outer-shell and the second (the major barrier liner) is in permanent contact with the fluid being transported.

Both of the polymers have the function of protecting the inner metallic layer from corrosion, by seawater and the fluid, respectively. The layer under study is the inner polymeric layer, which needs to have chemical and mechanical compatibility with the fluid being transported, in order to safely transport it.

There are three main polymers that are currently used for the inner layer: poly(vinylidene fluoride) (PVDF, crosslinked polyethylene (XLPE) and polyamide 11 (PA11). The choice of the polymer is determined by cost and by the operational conditions, such as temperature, pressure and fluid type. CO2 is likely to be in the supercritical state upon transport, because of the required temperature and pressure for storage and use.

Despite being non-toxic and non-flammable, the interaction of supercritical CO2 with the inner polymeric layer is a phenomenon of great importance for the pipeline stability. Under this stage, there are two main integrity challenges regarding the contact of supercritical CO2 with polymers: the swelling of the polymer, which can lead to rupture of the pipeline, and the gradual degradation of the polymer, that can lead to a loss of some key barrier properties of the polymer.

The removal of plasticizer from the polymer by supercritical CO2 may also be an unwanted effect in the case of PA-11 (which is the only polymer of the three mentioned above, and considered in this work, which contains plasticizer). In the design of flexible pipelines, the thermodynamic and transport properties, in particular the solubility, diffusion and permeability of the gas in the polymer, need to be carefully understood, since they determine how much gas escapes from the pipeline through the polymer barrier.

These properties vary with temperature, pressure, fluid composition and polymer type. The experimental study of these properties proved to be quite challenging, since the polymer physical properties, such as density, free-volume, volume, are dynamic, and so they change during operation at extreme conditions.

Furthermore high pressure/high temperature measurements – especially with CO2 – are never straightforward. Measurements were made at pressures up to 650 bar The permeability represents the overall mass transport across the membrane and accounts for the diffusion and solubility of the gas in the membrane (e.g. quantifies the amount of gas escaping to the metal confinement).

In this work, the solubility of pure CO2 is measured using a Magnetic Suspension Balance (MSB) for XLPE and for PVDF, for temperatures up to 403 K and pressures up to 300 bar. It is observed that the solubility temperature dependence followed the Arrhenius equation, decaying with temperature increase.

The solubility also increased with pressure. The experimental results for the solubility were modelled with the sPC-SAFT equation of state, which was able to correlate the experimental data; although a temperature dependent, binary interaction parameter was required. The polymer swelling is estimated based on an experimental method and using sPC-SAFT.

Modelling the swelling allows for different choices of the binary interaction parameters: It can be obtained from solubility in order to predict swelling or directly fitted to the experimental data. The polymer swelling increases with temperature for PVDF and decreases for XLPE, this effect might be due to the very high degree of crosslinking present in XLPE and not in PVDF.

A 2D-permeation cell is used to measure the permeability of pure CO2 and gas mixtures with high concentration of CO2 in PVDF, XLPE and PA11. The permeability of gas through PVDF was measured at pressures up to 345 bar and temperatures up to 403 K. In the case of XLPE and PA-11 the permeability is measured up to 650 bar, and at temperatures up to 363 K.

It is shown that the permeability always increases with increasing temperature, although the permeability increases with pressure for PVDF and PA-11 and decreases with pressure for XLPE. This trend is explained by the contrary effects that the pressure has on the free-volume, which may decrease because of the increase in the polymer density, or increase due to the penetrant increase that can lead to the plasticization of the polymeric chains.

The only plasticized polymer in the studied set of polymers (PA11), shows a loss in weight, from the pre to post-test, on average by 2.58% loss by weight. This effect is not observed in the other polymers, so it is assumed that the plasticizer is being removed from the polymer. Besides pure CO2, gas mixtures were also studied: several measurements were made for the mixture 90 mol% CO2 + 10 mol% CH4, although other concentrations of this mixture were also considered.

The presence of CH4 tends to decrease the total gas permeability, as expected, since CH4 is less permeable than CO2. In the other gas concentrations studied, it was difficult to draw clear conclusions regarding the effect of CO2 gas concentration, since only two pressures were studied and both show different behaviour.

The selectivity of the permeation through the membrane was analysed by gas chromatography. For XLPE the initial concentration was maintained (i.e. there was no clear selectivity for CO2), whereas PVDF was more permeable to CO2 than CH4, meaning that at the end of the experiments the concentration of CO2 increase compared to the initial concentration.

With the measured solubility and permeability, the diffusion was calculated. It shows the diffusion has a more pronounced effect in the permeability than the solubility, with respect both to temperature and pressure dependence; with increasing pressure, diffusion increases in PDVF and decreases in XLPE, while solubility increases in both polymers, the permeability shows the same pressure behaviour as diffusion, opposite to the solubility.