CHOOSE LANGUAGE:

Change Language Deutsch English Español Français Italiano Nederland Norsk Português 中文

• Print Page
• Bookmark Page

Oil & Gas

Although the rapid gas decompression (RGD) resistance test regimes generated by major operators — such as Total and Shell — are highly regarded and widely accepted, the independent qualification of choice for many equipment suppliers and fluid sealing specialists is Norsok M-710 Annex B.

Developed by the Norwegian petroleum industry as a method for the qualification of non-metallic sealing materials and manufacturers, Norsok M-710 Rev 2 (October 2001) defines the requirements for critical sealing, seat and back-up materials for permanent subsea use, and is also applied to topside valves in critical gas systems. Annex A of the standard concerns the ageing of elastomeric materials, whereas Annex B covers the RGD testing of elastomeric materials. Annex C concentrates on the ageing of thermoplastics.

The complete Norsok M-710 is currently being updated and converted to ISO 23936-2: Petroleum, petrochemical and natural gas industries — Non-metallic materials in contact with media related to oil and gas production — Part 2: Elastomers. An industry-based committee with representation from James Walker and other seal manufacturers is performing this task. ISO 23936-2 is likely to be published in late 2011.

Norsok in a nutshell

The Norsok M-710 RGD test regime requires a minimum of three BS 1806 ‘O’ rings of 5.33mm section and 37.47mm ID to be inserted in housing grooves and subjected to a pressure of either 150bar, 200bar or 300bar (2175psi, 2900psi or 4351psi) at temperatures of 100°C, 150°C or 200°C (212°F, 302°F or 392°F). For both sweet and sour wells, the fluid media is either 3% CO₂ + 97% CH₄ (‘low’ CO₂), or 10% CO₂ + 90% CH₄.(‘high’ CO₂). For carbon-dioxide injection wells, 100% CO₂ is used.

After heating the test rig with samples in situ, pressure is applied for 72 hours to allow the media to permeate into the ‘O’ ring. Decompression to ambient pressure is then undertaken at a rate of between 20bar and 40bar (290psi and 580psi) per minute. The sample is held for one hour at ambient, then re-pressurised and soaked at temperature and pressure for 24 hours before further decompression. This cycle is repeated a total of ten times before the test rig is allowed to cool to ambient temperature and left for 24 hours before removing the ‘O’ rings.

Each ‘O’ ring is cut into four equal radial sections and visually examined under magnification of at least x10. The RGD damage at each section is rated between 0 and 5, according to the following criteria:

0 rating: No internal cracks, holes or blisters of any size.

1 rating: Less than four internal cracks, each shorter than 50% of cross section, with a total crack length less than the cross section.

2 rating: Less than six internal cracks, each shorter than 50% of cross section, with a total crack length less than 2.5 times the cross section.

3 rating: Less than nine internal cracks, of which a maximum of two cracks can have a length between 50% and 80% of the cross section.

4 rating: More than eight internal cracks, or one or more cracks longer than 80% of the cross section.

5 rating: One or more cracks going through the cross section, or complete separation of the seal into fragments.

Norsok stipulates that ratings of 4 and 5 are not acceptable (‘fail’), which means that ratings of 0, 1, 2 and 3, can be considered as acceptable (‘pass’) for RGD resistance.

0 rating - Perfect Pass	1 rating - Pass	2 rating - Pass
3 rating - Pass	4 rating - Fail	5 rating - Fail

Each ring’s performance is defined by its rating on all four of the cut sections. Thus a 0000 rating defines a ring with a perfect ‘pass’ and no visible internal faults, whereas a 3434 rating would be defined as a ‘failure’. The highest numbers (ie, poorest performances) are taken for the three samples tested at the same time to give an overall rating for the elastomer under the stated test conditions.

Perfect pass or compromised safety?

At James Walker, we test four samples instead of the minimum of three, and use the 0000 rating (zero damage across four cross sections on all four samples) as a benchmark when developing our RGD-resistant materials. Our reasoning is that a 0000 rating represents the maximum achievable operational safety margin for end users — this is especially valid for long-term applications where fluctuations in media composition and temperature can further compromise RGD performance.

However, it is always possible that a minor surface blister may form during the test and, when the seal is sectioned, this would be recorded as a 1 rating. For this reason, it is imperative that photographs of the seal cross sections are included in any report, whether for Norsok or other accreditations.

As noted above, it is possible to claim a ‘pass’ against Norsok M-710 Annex B with a rating of 3333. As the pictures (above) show, the potential difference between a 3 rating (‘pass’) and a 4 rating (‘fail’) can be very fine. Therefore, under true operational conditions, damage at the 3 rating level offers the user virtually no safety margin. In our experience, such a material is unfit for RGD-resistant duties.

James Walker’s vast experience in this field dates back to the late 1970s and early 1980s, when we initiated a comprehensive research programme into the phenomenon of RGD (then known as ED or explosive decompression) following reports of seal failures in North Sea fields during exploration and production activities. In collaboration with the research arms of two offshore operators, and oilfield equipment manufacturers, we investigated the causes of RGD damage with gas permeation and diffusion into elastomeric components at very high pressures.

We have been continuing this research for 30-plus years at James Walker Technology Centre, where we have some of the world’s most advanced RGD testing facilities that match or surpass those used by most independent test houses. Currently we test ‘O’ rings up to 10mm cross section within the ranges of 150bar (2175psi) at 200°C (392°F), to 350bar (5076psi) at 65°C (149°F), with full data-logging of temperature and pressure.

Understanding RGD processes

Our current understanding is that an RGD event occurs as two distinct processes: permeation, and pressure release — with possible phase changes in the fluid media. 

Under pressure, a fluid passes through the elastomer surface and towards the core of a seal, dissolving into the material. Even at the very high service pressures specified by some end users, such as 690bar to 1276bar (10,000psi to 18,500psi), the fluid is generally in a gaseous phase. However, at high pressures and low temperature, some chemicals/mixtures can be present as supercritical fluids — where the liquid and gas phases approach each other. 

The fluid media very slowly saturates the elastomer (extremely slowly at supercritical levels) to equalise the pressure between the seal surface and its core. Complete saturation can take a considerable time — typically days or weeks, rather than hours. 

Interestingly, this lengthy permeation process led to some unreliable test results in the very early days of developing RGD-resistant elastomers, as the test regimes used at that stage did not allow sufficient time for complete saturation of the material down to core level. 

When external pressure is suddenly released — such as a blow down — the media within the seal expands very rapidly in its gaseous state, accompanied by a temperature drop caused by adiabatic expansion. The gas expands much faster than it can naturally diffuse through the elastomer. If the elastomer cannot resist crack or blister growth under these rapid decompression conditions, then the seal will suffer structural failure. 

The survivability of an elastomeric seal during an RGD event is also dependent on the material’s resistance to degradation in oilfield media over long-term service, as the material’s mechanical properties will reduce with saturation time.

Elastomer ageing estimates

Norsok M-710 Annex A specifies procedures for conducting accelerated ageing tests in an autoclave on constrained ‘O’ ring samples at 100bar (1450psi) or higher, and at three or more different elevated temperatures — all of which must be above the service temperature. The constituents of test media are defined for sweet and sour service conditions, albeit other test fluids may be used to match specific operating conditions. 

The mechanical properties of the ‘O’ ring samples are measured before and after the accelerated test, with acceptance criteria based on percentage swell, and changes in material hardness, tensile strength and E-modulus. Service life of the material is estimated by extrapolation of the tensile properties using the Arrhenius method.

James Walker’s experience of this method for ageing elastomeric materials shows that some highly unrealistic values can be achieved. We would like to believe that our FR58/90 RGD-resistant fluorocarbon has a service life in sour gas of 817 years before its tensile strength is halved — as determined by a leading independent test house. However, we must accept that the results of Norsok’s Annex A ageing procedure can be used only as a comparison between different materials; not as an accurate guide to service life expectancy. 

We have discovered two additional factors that potentially cause material degradation during the Norsok ageing test, in addition to the effects of sour gas, which are the basis of the test procedure. These are RGD events and the effect of solvents, both of which can change the physical properties of the test samples — either independently and/or in concert with the effect of sour gas — thus significantly affecting the estimated service life prediction. 

Firstly, in some instances when a sample is removed from its housing after a sour ageing test, it pops and crackles as RGD events occur. This action can significantly affect the material’s tensile strength. Secondly, the stipulated carrier (at 60% by volume of total test fluid) for the sour oilfield media is a cocktail of heptane, cyclo-hexane and toluene. These cause swelling and the reduction of tensile strength and other properties in a number of RGD-resistant elastomers.

Designing for RGD resistance

Intimate knowledge of the characteristics of different elastomers is vital when designing and specifying high-integrity sealing products for service in RGD-related applications. 

For example, a hydrogenated nitrile (HNBR) compound, such as Elast-O-Lion® 101, has sufficient tensile strength to withstand an RGD event, and can operate constantly at up to 160°C (320°F). A fluorocarbon (FKM) compound, such as FR25/90, has the higher constant service temperature of 200°C (392°F). However, FKM is physically weaker than HNBR at elevated temperatures. 

We learned many years ago that at elevated temperatures and pressures, significant engineering expertise must be introduced to enable the specific benefits of an RGD-resistant elastomer to be fully realised. 

One of the major aspects is the design of the housing in which the seal is enclosed. In our experience, the design of a new housing must be undertaken by experienced sealing engineers who are fully conversant with the optimisation of an RGD-resistant system. Simply employing a conventional ‘O’ seal engineering approach can severely compromise the effectiveness of an RGD-resistant seal. 

A further aspect is that RGD-resistance is influenced considerably by the cross-section of the seal — the thicker the seal, the more difficult it is to resist RGD. At 5.33mm cross-section, many specialised RGD elastomers will gain close to a 0000 rating. 

We have recently achieved success with a 10mm section ‘O’ ring of our new generation FR68/90 fluorocarbon in tests at 100°C (212°F), and a decompression rate of 35bar (508psi) per minute. This high decompression rate induces significantly more internal stress in the ‘O’ rings, and more accurately represents an operational RGD event than a decompression rate at the lower end of the Norsok range..

The next challenge is to achieve a perfect Norsok rating with a 12.7mm cross-section ‘O’ ring under realistic service conditions. This means using an elastomer, such as FR68/90 that is optimised for sealing efficiency as well as RGD resistance and does not need an exacting surface finish on the counterface, or split housing construction to overcome compound elasticity limitations, for high-integrity sealing.

Manufacturing for RGD resistance

We consider that the effective compounding of an RGD-resistant elastomer is the most important part of a seal’s manufacture. In-house compounding enables the manufacturer to exert total control over variables such as raw ingredients, temperature, time and the shear energy imposed on the compound by the mixer’s rotors. 

We optimise our ingredients and mix cycles to achieve precisely the right compound for the RGD and sealing qualities required, then always mix the compound on the same machine under exacting computer control for total repeatability and an homogeneous mix. High levels of dispersion and consistent heat history are critical and necessary to produce the quality required. 

James Walker always uses the same materials from the same suppliers that were initially used for Norsok or other accreditation testing. To ensure consistency of product, any potential ingredient change (including a change in manufacturing plant) involves revalidating the product prior to acceptance.

Putting out the compounding to a contractor imposes the potential risk of inconsistency in the resulting elastomer. Using different mixers and/or different mix cycles will produce very different compounds. 

The correct moulding and post-moulding techniques used to convert an elastomer into an efficient seal is critical to obtain RGD performance, and must be optimised for each material. Again, this needs to be an in-house operation to achieve the consistency of results that are required. 

A manufacturer needs to know on which particular presses, and at which computer-controlled settings, a seal has been produced in the past to achieve the required results. Unacceptable inconsistencies can result from shifting the moulding process to another type of press, or even to an identical machine at another plant. We routinely take random samples from production runs and test these in-house under Norsok M-710 Annex B, or other accreditation requirements, to ensure product quality. 

The lesson here is once you have set a process cycle that achieves perfect results for RGD-resistant seals, do not change any parameters even by the smallest amount. Otherwise, you will need to re-establish the complete process from point zero to maintain product standards, and that can prove a time-consuming and expensive remedy. 

Total in-house control of every variable is vital for RGD-resistance with elastomeric seals. At James Walker, we adhere strictly to this principle to achieve the very best quality RGD-resistant seals for our customers. 

Click here for a pdf copy of our guide to Elastomeric seals & components for the Oil & Gas Industry.