Fluid movements through polymers...
TWI Bulletin, July - August 2009
...permeation and diffusion under the spotlight
Richard Shepherd joined TWI in the Spring of 2007. For the 25 years prior to this he has worked for various R&D and engineering organisations, working with engineering applications for polymers; many in the oil & gas sector. He has published a number of papers specifically in the use of polymers in oil field applications and is joint inventor on a patent for high pressure and temperature dynamic seals. Richard has qualifications in Mechanical and Mechanical and Production Engineering. Currently he is working in a number of areas including polymers in sweet and sour environments, permeation of polymers, pipe coatings, flexible joint technologies and non-destructive test methods for high value polymeric components.
Aurelie Brun joined TWI in October 2006 as a project leader. She is currently working in the polymer section and is mainly involved in laser processes of polymers and permeation measurements of gases through polymers using gas chromatography.
With the growing use of polymers in all sectors comes the exposure of these materials to increasingly hostile fluids. Richard Shepherd and Aurelie Brun consider here some recent developments in testing and qualification methods available at TWI, focusing on the phenomena of diffusion and permeation, in particular for the automotive and oil and gas sector.
Permeation and associated phenomena
The initial driving force for permeation through a polymer is the chemical potential of the fluid contacting it. As a result, permeation occurs by permeant molecules entering it, migrating through its bulk via the finely-divided internal space which exists between the polymer's own molecules, and escaping from the far surface (Fig.1). Therefore it is possible to consider the mechanism of permeation mathematically from the combination of the same three stages of fluid transmission; into, through, and out of the polymer.
Fig.1. Permeation phenomena through polymer membranes
In the first stage the permeant dissolves into the surface of the material (adsorption). The second stage is the diffusion of the permeant molecules through the material bulk under the action of the concentration gradient (ie from high concentration to low), according to Fick's laws. The final stage is evaporation of the permeant from the low pressure surface out into the adjacent atmosphere. It is not the intention to develop the various mathematical equations describing these phenomena here, rather to describe the application of diffusion and permeation measurements currently available and under development at TWI.
Automotive
The fuel system components of vehicles are often made from various types of polymeric materials. Since no polymeric material forms a complete barrier to the passage of fluid molecules, these components are a source of small but continuous evaporative emissions. Even if all probabilities of leakages are removed, the molecular phenomenon of permeation means that a small but definite amount of the fluid will eventually migrate through the material.
Legislation sets 'near-zero' limits for evaporative emission globally (EURO II, III/2000 and CARB legislation). This legislation represents a major challenge for the use of polymers in automotive fuel systems. Hence, where polymers are used, reduction of fuel permeation is a major requirement in bringing about reductions in overall emissions to acceptable and legislative levels. Rather paradoxically, as materials with improved permeation resistance are developed, so the difficulty of measuring their permeation rate increases.
Oil and gas
Knowledge of high pressure gas, and indeed liquid, permeation rates is essential for designing any polymeric component for use as a seal or transport system in the oil and gas sector. Excessive permeation can lead to so-called fugitive emissions with resultant corrosion or worse Health and Safety issues. TWI has invested significantly to enhance further the testing methods to characterise the polymer materials used in the manufacture and construction of thermoplastic-based flexible pipes for offshore oil and gas production, see Figure 2.
Fig.2. Typical structure of unbonded thermoplastic flexible pipe as used in the oil and gas sector
As previously described, gases can dissolve into the surface of a polymer (adsorption) and then diffuse into its interior. If geometrical and/or surrounding considerations allow, the gases will pass right through the elastomer by dissolving, diffusing and evaporating, the combination of these processes being permeation. In addition, water from the sea could penetrate the polymers. Again, as described, these phenomena are characterised for a material by the permeation coefficient (Q) - a product of the diffusion coefficient D and solubility parameters. The diffusion rates of gases are approximately 1-2 orders of magnitude faster than those for liquid diffusion.
Liquid permeation testing
For liquid permeation, the Thwing-Albert transmission cup mass loss test provides a direct and simple measurement of its rate (Fig.3). However, possible disadvantages of this technique are that problems in achieving an effective seal may occur, and long test periods may be required before equilibrium is reached. Moreover, only one test is possible per cup. In contrast, total immersion absorption tests, which are described below and can give reliable estimations of permeation rates, require no sealing. They have the advantage that several specimens can be immersed in the same container of fuel, sheet specimens of any shape can be used and tests can be somewhat shorter than those for permeation.
Fig.3. Schematic of Thwing Albert transmission cup
Transmission cup method
The transmission cup or Thwing-Albert cup equipment (Fig.3) and associated procedures detailed by standard ISO 6179 (BS 903:A46), which applies for volatile liquids, can be used for permeation tests. The test fluid (fuel or other volatile fluid) is sealed into the cup by the clamped test specimen sheet, and the loss of weight of the whole assembly is obtained at intervals and plotted.
Initially a certain time elapses before steady state permeation is achieved. This transient stage provides an opportunity to determine D the diffusion coefficient for the material (Fig.4).
Fig.4. Q and D obtaied from a permeation rate plot
Using this method also allows for the derivation of Q, the permeation coefficient of the material (where P1 is the vapour pressure through the sheet specimen of area A, and P2 is a region of lower vapour pressure, the surface exposed to the atmosphere).
The equation above demonstrates the simple but important observation that permeation rate is directly proportional to cross-sectional area and inversely proportional to thickness. Hence predictions can be made via this equation of permeation rates for sheets of different dimensions for the same fluid/polymer system.
Absorption tests
Absorption testing in a liquid is a simple process as indicated in Figure 4. The sheet sample is removed periodically, rapidly weighed, and replaced.
As for the previous method, predictions can be made via this equation of permeation rates for sheets of different dimensions for the same fluid/polymer system.
The behaviour shown in Fig.5 conforms to Fick's Law which enables the calculation of permeation and diffusion coefficients.
Fig.5. Immersion testing method
Acceleration methods
Clearly at room temperature or below diffusion rates can be relatively low. Acceleration of these processes is possible by the increase of temperature within certain materials limits such as Tg. The use of three or more accelerating temperatures allows the development of an Arrhenius relationship for that material and fluid allowing extrapolation of permeation behaviour back to lower temperatures. TWI has undertaken a number of materials characterisation studies using absorption methods to determine diffusion rates of materials systems and by combining these methods with evaluation of mechanical properties can inform the end-user with respect to durability.
Water permeation using dew point measurements
TWI has invested in dew point measurement equipment to measure the amount of ppm of water permeating through the polymer membrane. The sample is placed in a permeation cell. A small amount of water is positioned on top of the polymer sample and high pressure N2 is flushed to vaporise the water. Moisture in the N2 flow is measured by the cooled mirror dew point measurement equipment. The equipment sensitivity is 0.2ppm.
Because it is difficult to guarantee that there will not already be moisture present in the system and in particular the sample, the graph obtained will not start at zero ppm of water. This will not allow measuring the diffusion and solubility coefficient. Absorption tests still have to be carried out for this.
Gas permeation testing
Manometric and volumetric methods
These methods are described in ASTM D1434-82 (Re-approved 2003). The polymer membrane to be tested is mounted in a gas transmission cell. It creates a seal between a chamber containing the test gas at high pressure and a chamber kept at constant pressure or volume condition. The change in volume (volumetric method) or pressure (manometric method) in the second chamber is monitored during the whole duration of the test. A graph of permeated gas pressure (or volume) against time is obtained (Fig.6).
Fig.6. Graph obtained for the manometric method
This allows measuring the diffusion coefficient as explained in the liquid permeation case, and from there the permeation coefficient can be deducted. The main disadvantage of this method is the fact that gas mixture can not be studied as a function of each of the gas composing the mixture.
Gas chromatography method
More recently at TWI, gas chromatographs have been used for these measurements. TWI has invested in sensitive gas chromatographs, which have been installed in TWI's unique H2S facilities. The equipment is capable of detecting minimum levels of less than 1ppm of CH4 and around 100ppb of H2S. They can detect CH4, CO2 and H2S. Both pure gases and gas mixtures can be studied. In addition, TWI uses a specially designed permeation cell, as shown in Fig.7.
The experiments are carried out so that the permeated gas is constantly flushed to the GC. This avoids accumulation of the test gas so that the amount measured is really what is going through the polymer sample at a certain instant in time. Figure 8 shows a typical graph obtained with this technique.
Fig.8. Permeation graph using the gas chromatograph method
| Q: permeation coefficient
D: diffusion coefficient
S: solubility coefficient
Fpermeated: flow of permeated gas
h: thickness of sample
A: surface of permeated area
P: pressure of gas test
t83%: time at which the flow of permeated gas is equal to 83% of the maximum flow of gas permeated |
TWI is currently undertaking an extensive range of permeation testing of polymers for the oil and gas industry. The work includes study of the permeation of pure CO2, CH4 and H2S, as well as a mix of these gases and oil.
The use of numerical modelling with diffusion data
The permeation, diffusion and solubility coefficients obtained experimentally can also be used to model gas permeation through materials. It is then possible to obtain an anticipated permeation rate as a function of time at the exit. An example is shown below modelling the permeation through an o-ring. To model the permeation realistically, the deformed geometry of the o-ring, as a result of the applied gas pressure, is first modelled (Fig.9).
Fig.9. Modelling of the mechanical deformation of the o-ring
The permeation rate near the exit is then calculated from the concentration flux and the area available for evaporation of the gas. Contours of mass concentration are obtained, as well as the predicted permeation rate of the gas through the o-ring as a function of time (Fig.10).
Fig.10a) Concentration contours for CO2 at equilibrium - zero concentration is indicated at the exit since this is where the gas evaporates following permeation
Fig.10b) Predicted CO2 permeation as a function of time
The plan
The current gas permeation infrastructure is capable of operating up to 200bar with the permeation cells rated to 1000bar and 200°C. With significant interest in carbon capture and sequestration, there are plans to upgrade the equipment so that pressures of 300bar can be accommodated to allow the study of the permeability of supercritical CO2 and the behaviour of materials in contact with the fluid. Armed with accurately measured permeation coefficients, it is possible to model the components numerically for their permeation behaviour and the subsequent corrosion environment that may or may not be created as a result.
Specifically for the automotive sector there are preliminary plans, subject to Member interest, to develop a flow loop type test facility to complement the current static type permeation characterisation testing described here to qualify fuel transmission systems to SAE standards for the latest generation of bio-fuels.
