For material of this hardness damage can be expected to occur with this amount of H2S in very short times. Unless temperatures are above 150C, you're asking for trouble.

I concur with mcguire.  We perform many tests on sulfide stress cracking resistance of carbon steels, and your material and hardness would be expected to fail within a very short period (hours, not days).

If you want an example I had improperly stress relieved 4130 (HRC 25 to 27 in the HAZ) that failed after 3 hours exposed to 2% H2S at 800 psi.

Hush,

Your component most likely failed due to chemical inhomogeneity inherent in solidification microstructure, not hardness of HRC 25-27.  If you have a homogeneous 4130 component with HRC 25-27 exposed to your conditions, it would last longer than 3 hours.  This is not to suggest that it would be suitable for an actual application - the desired hardness would be around HRC 23.

As an aside, a more appropriate measure of resistance is fracture toughness, e.g. KIssc measured according to NACE TM0177.

CoryPad,

Could you further explain the concept of chemical inhomogeneity in solidification microstructure? Keep in mind I'm not a metallurgist. Is this inherent to some degree in all welds or is it due to weld technique?

For background, the failure in question was in the parent material (4130 Normalized, Q&T, HRC 22 max) approximately 3/16" from the fusion zone. Cracks propogated radially from a machining mark (1/32" deep sharp scratch)in the bore. Hardness traverse across sectioned piece showed HRC of 19 in parent material climbing to HRC 27 in HAZ dropping to HRC 25 in weld. Area had high stresses in tension as well as the pressure load. Retained fluid was wet gas, some chlorides & approximately 2% H2S at 800 to 1000 psi and about 150 F.

I realize there would have been a slight stress riser at the machining mark but could failure possibly have been promoted by inclusions in the 4130? (I'm thinking the machining mark could have been caused by a hard inclusion).

Hush

I'll let CoryPad give you his official answer, but in one word...bingo.

Certainly, Hush.  Upon solidifying, a liquid solution will partition.  The solid solvent will retain as much solute as possible, and the rest will precipitate as a second solid.  For iron-carbon alloys, let's simplify the matter and call iron the solvent and iron carbide the solute.

Continuing with the simplification, when welding you have liquid iron/iron carbide solution.  When it solidifies, you will have solid iron/iron carbide solution + solid iron carbide.  The shape, concentration, and distribution of the carbide is heterogeneous.  The heterogeneity is bad for mechanical and corrosion properties.  This is a condition for all as-solidified material: cast iron, cast steel, welds, etc.  This is the reason one should heat treat welds - it homogenizes the microstructure.  

In addition to carbide structure, one must also contend with inclusions.  They also will be heterogeneous directly after solidification.

Another complication is residual stress.  This routinely is present after solidification.

One last complication is microstructure in Heat Affected Zones (HAZ) of the joined parts.  The carbide microstructure may change due to heat from welding, even though the HAZ may never have liquified.

So, in my original post, I did not mention the HAZ, residual stress, or inclusion factors.  These, in addition to the poor microstructure present in as-solidified welds, are the root cause of short-time sulfide stress cracking failures, not a hardness change from 22 HRC to 27 HRC.  This is not to suggest that 27 HRC will provide a useful lifetime - it may or may not.

CoryPad,

Many thanks for the info/lesson. Your explanation has given me some new insight into possible failure mechanisms. If you get a chance should build this into an FAQ, I'm sure others would benefit.

PaulEng, my apologies for hijacking your thread.

 

This is one of the highest level discussions I have seen on  eng-tips.com. Hush, CoryPad, and TVP are very knowledgable materials engineers. This is a exceptional forum.
 I suggest that the HAZ structure with its HRC 27 and untempered nature is sufficent for susceptibility. Residual stress is nasty, and stress raisers are worse.
 It's natural to suspect inclusions, but this is very unlikely. Welds are cleaner than the parent metal and the parent metal is generally very clean for alloy steels.
 You simply must get the hardness down to resist hydrogen embrittlement. Toughness counts next, and untempered structures are generally lacking in this regard unless carbon level is very low.

No problem Hush - it's been a good learning experience/discussion for me too!

mcguire is being modest for not including himself/herself in the list of excellent contributors to this thread (TipMaster of the week for this week!).  Anyway, you may be surprised at how much influence inclusions can have on SSC behavior.  I am familiar with the testing that CoryPad performs, and there is some good data to show that for a given level of strength/hardness, the presence of higher levels of inclusions and interstitial atoms (oxygen and nitrogen) can result in a noticeable decrease in SSC resistance.  The steel grades being tested are all very clean alloy steels that have been processed using vacuum degassing, but there is still a significant difference due to microstructure-- solidification structure, grain size, inclusion content, gas content (O & N), etc.

Based on the further discussion, it sounds like the residual stresses and untempered structure are more likely to have been the first order effects for Hush's situation.  However, toughness tracks very closely with SSC performance, so anything that reduces fracture toughness will definitely reduce SSC resistance.