Rapid gas decompression looks like a material that has been cooked. The ring comes out of a joint that held perfectly for months, and it is blistered, spongy, or cracked through the section — sometimes still intact on the outside and crumbly when cut. Nothing in the process temperature explains it, because the temperature did not do it. The pressure release did.

The mechanism
Under high gas pressure, gas dissolves into the elastomer until the concentration reaches equilibrium with the pressure. How much dissolves depends on the gas and the compound: CO₂ dissolves far more than nitrogen, hydrogen more than helium by volume, and a high-fluorine compound less than a hydrocarbon one.
When the pressure is released, the dissolved gas has to diffuse out. Diffusion is slow — minutes to hours through a section a few millimetres thick. If the external pressure falls faster than the gas can escape, the internal pressure exceeds the external, and the gas comes out of solution inside the rubber as bubbles. If the resulting pressure exceeds the tensile strength of the compound, the bubbles grow into voids and cracks.
Two thresholds govern whether damage occurs:
- Pressure level. Below roughly 1,000 psi (70 bar) the dissolved volume is usually too small to cause damage. Above 3,000 psi (200 bar) the risk becomes significant for most compounds.
- Decompression rate. A slow, controlled bleed lets the gas diffuse out without exceeding the internal strength. A sudden release — a valve opening to atmosphere — does not.
Temperature works against you twice: it increases the amount of gas dissolved and it softens the compound, lowering the pressure the bubbles need to cause damage.
Why it appears after shutdown
A joint in continuous service at steady pressure has no decompression event. The damage happens when the system is shut down, depressurised for maintenance, or hits a pressure excursion with a fast release. This is why RGD failures are reported as "it was fine until we took it apart" — the seal was fine, and the disassembly destroyed it.
The practical consequence: a ring removed from gas service cannot be judged by appearance alone. If the section is spongy when cut, or shows voids under a lens, the ring is scrap regardless of how it looks outside.
Compounds that resist it
RGD resistance is not the same property as chemical resistance, and a compound specified for the media may still fail on decompression. What matters:
- High hardness. 90 Shore A resists bubble growth better than 70 A, because the compound has less free volume and higher strength.
- High crosslink density. Same reason: a tighter network resists void growth.
- Low gas permeability. Reduces the amount dissolved in the first place.
- Low swell. A swollen compound has more free volume and loses strength.
In practice, dedicated RGD grades exist for FKM, HNBR and FFKM, tested to NORSOK M-710 or ISO 23936-2. A general-purpose FKM that passes a media compatibility test may still fail an RGD test, which is why the two specifications are separate.
| Service | Typical choice |
|---|---|
| CO₂ and natural gas, moderate pressure | 90A HNBR or RGD-grade FKM |
| High-pressure gas with aromatics | RGD-grade FKM, or FFKM where temperature is also high |
| Hydrogen | Compounds tested for hydrogen service; permeability and RGD both matter |
| SCUBA and breathing air | 90A NBR or EPDM, with proven decompression behaviour |
Changing the decompression profile
The cheapest fix is often procedural, not a material change:
- Limit the depressurisation rate. Many RGD specifications give a maximum rate — often in the range of 20–70 bar per minute, depending on the standard and the compound. A controlled bleed valve costs far less than an FFKM seal.
- Step the pressure down. Hold at an intermediate pressure for a period to let gas diffuse before the final release. The hold time depends on section thickness and gas; a few minutes at 30–50 % of working pressure captures much of the benefit.
- Avoid trapping gas. A groove design that traps a volume of gas against the ring creates a local decompression event even when the system pressure falls slowly.
- Reduce section thickness where possible. Diffusion time scales with the square of the thickness, so a thinner section degrades faster — for RGD, thin is better.
The same ring survives in one installation and fails in another at the same pressure, because the decompression profile is different. If you cannot control the profile, specify a compound with margin.
Distinguishing RGD from other damage
| Appearance | Likely cause |
|---|---|
| Blisters, voids, spongy section | RGD |
| Cracks on the surface, hard ring | Thermal ageing |
| Swollen, soft, solvent smell | Chemical attack |
| Chewed edge on the low-pressure side | Extrusion |
| Clean cut, faces match | Installation damage |
The give-away for RGD is that the damage is volumetric — inside the section, not on a surface. Cut the ring and look at the cross-section: thermal ageing and chemical attack show on the surface, RGD shows through the body.
Where this is heading: hydrogen
Hydrogen service is the fastest-growing reason to ask about RGD. Hydrogen molecules are small and diffuse quickly, so the decompression window is shorter than for the same pressure in nitrogen, and the consequences of a leak are different. Compounds are being qualified against hydrogen-specific protocols, and the practical advice is the same as elsewhere: specify a compound that has been tested for the gas and the decompression profile you actually have, not for gas service in general.
Frequently asked questions
At what pressure does RGD become a risk?
From roughly 1,000 psi (70 bar) upward for fast decompression, and above 3,000 psi (200 bar) for most compounds even at moderate rates. The rate matters as much as the pressure.
Can RGD damage be seen without cutting the ring?
Sometimes — blisters on the surface are visible. But a ring can be intact outside and voided inside, so cut a section from a suspect ring before concluding it is serviceable.
Is a harder compound always better for gas service?
For RGD resistance, generally yes, because 90A has less free volume and higher strength. But hardness also raises seating load and reduces conformability, so the joint has to suit the harder ring.
Does RGD happen in liquids?
Not in the same way. Liquids do not dissolve into elastomers to the same extent, and they do not expand explosively on depressurisation. The failure mode is specific to gases and to two-phase systems.
What standard should I specify?
NORSOK M-710 and ISO 23936-2 are the two most quoted for oil and gas service. Ask for the test conditions — gas, pressure, temperature, decompression rate — because a pass at one condition does not imply a pass at another.