Oil & Gas

Sour Gas & You

Jul. 19 2021

It happens a lot: we get a call that says “I’ve got a string that was exposed to sour gas, so I’m going to send you one joint so you can test it to see if it’s embrittled.”

Deep breath.

There are multiple exciting things happening here, and I have a hard time straightening them out. In just those two sentences from my imaginary customer-to-be(-or-maybe-not) I’ve got four problems:

  • “…was exposed to sour gas …”
  • “… send you one joint …”
  • “… test it to see if it’s embrittled …”
  • “… so I can get my money out of those rascals …” (that’s implied—but obvious)

Let’s see if a little science can straighten things out. First, “sour gas” is hydrogen sulfide (H2S, or hydrogen sulphide for our British friends), and it’s super-nasty stuff. HSE issues aside (really, though, it’ll kill you dead, don’t mess around), it’s also not good for our pipe because of this:

    H2S + Fe => FeS + 2H+

If your memories of high-school chemistry mainly involve Jenny Smith and her pink sweater, let’s do that in words. (She married a dentist and moved to Sacramento—try to focus.) That equation means that hydrogen sulfide combines with iron (which is the main component in our steel pipe, clearly) to make iron sulfide and hydrogen. 
 

Image
Blue Diatomic Molecule

To remember a little more, hydrogen is usually H2, that is, a “diatomic molecule” made up of two hydrogen atoms stuck together. Here, though, it’s actually two individual instances of hydrogen atoms—essentially just a proton bopping around. An H2 molecule is big enough to bounce off steel, as illustrated by the fact that you can contain hydrogen in a steel tank.

Image
H2 Molecule

That ionic hydrogen atom, on the other hand, can drift right through the steel. When it’s inside the steel, it tends to gather at grain boundaries because those are high-energy locations. With tensile stress applied, the ionic hydrogen will cause the steel to crack and fail along those grain boundaries.

When the cracking and failure happens, it’s a brittle fracture. (You remember brittle versus ductile, right? Ductile materials stretch and deform before they fail; brittle materials just pop apart.) But the steel wasn’t brittle—it was good, ductile stuff to begin with. Sulfide Stress Cracking (SSC, which is what we’re talking about here) makes a good, ductile material fail in a bad, brittle way.

So, to deal with my problems above:

“…was exposed to sour gas …”

This doesn’t really mean anything because there are degrees of “exposure” that depend entirely upon a whole host of both environmental and material factors. For instance:

  • If the temperature was hot, the hydrogen ions are as energetic as Labrador puppy, meaning that most of them run into each other randomly and couple into H2. The hydrogen gas then floats harmlessly away before it gets a chance to get into the pipe. This is backwards from most corrosion mechanisms—most of the time high temperatures are bad, but here it’s good.
  • If the pH was high, the OH- ions suck up the H+ ions and work to slow the cracking reaction, and vice versa.
  • In oil-based mud the reaction just doesn’t happen, so it’s much safer. (Maybe; did you know some oil-based mud has water in it?)
  • Higher hardness materials are much more likely to crack than low-hardness materials.

I could go on; the point is that exposure to sour gas is not monolithic, and whether or not you should be worried depends on a lot of other stuff.

“… send you one joint …”

I have this problem with not just SSC, but any form of inspection. Imagine your string was exposed in such a way that 5% of the joints in that string are now cracked or otherwise problematic. If you randomly choose one joint, you have a 1 in 20 chance of getting a bad one. However, if you put the string back in service, you have about a 20 in 20 chance of having a failure when the weak link gives way. Sampling can be useful in other areas (manufacturing, for instance), but not this. If you need to do an inspection, you need to inspect everything.

“… test it to see if it’s embrittled …”

Sour gas makes good steel fail in a brittle way; it does not turn good steel into bad steel. If the pipe is not actively cracked, then it will be just as good today as it was when it was made.

I know that’s not what you thought; this is where our folk wisdom fails us. The hydrogen ions gather around the grain boundaries and make them susceptible to failure; if there is no failure (cracking) and the hydrogen ions go away (which they will once you take them out of the sour environment), there’s nothing left to damage your pipe. My laboratory testing won’t tell you anything.

“… so I can get my money out of those rascals …”

Because of this, any question that someone asks me is bound to be loaded. You’ll likely quote me the next time you’re arguing about this with the other guy, but you’ll only pick the parts you like, and so will he, so my actual opinion will get mangled like a hog through a grinder. But here goes.

In theory, if you inspect the whole string for cracking, and you find and remove any joint that has cracking, then what’s left is ready for use. Sour gas causes cracking, but if it doesn’t get a chance to cause cracking it doesn’t matter for the future.

However, there is no such thing as a perfect inspection, not even when you use DS-1 and a third-party monitor. We do the very best we can, but our instruments and processes are not sensitive enough to guarantee you’ve found every possible crack. (We can get really close, but I can’t make an absolute promise of success.) As such, an “exposed” string does have an elevated risk of failure, small though it may be.

So what do we do? I say inspect it thoroughly (using something that’s good at finding small surface cracks—MPI with an AC yoke and dry powder is great, EMI is usually more reasonable for full-length joints). Reject anything with cracking (of course) and put the rest back in service. If your risk tolerance is super low, inspect it again pretty quickly just in case some operational time opened up cracks you might have missed the first time.

Which means that, if the guy on the phone believes me … I just talked myself out of some business. C’est la vie.

Image
Grant Pettit

Grant has been at BVNA for 15 years, doing failure analysis, drill stem and casing design, standards writing, and teaching others to do it, too.

Grant PettitDirector Of Operations - Standards, Training, & Accreditation