SS vs. CS in sour gas service 2

[Tenoa]

I have been looking at the NACE MR0175 / ISO 15156-1 standard and trying to determine the discharge material requirement for a sour gas compressor skid at 5000psig, with up to 10 mol% H2S present. The standard is very detailed in it's description of allowable carbon steel and stainless steel material compositions and requirements, but it does not state when each should be used. I know the norm is SS for acid gas systems, but I don't know what makes it "acid" or why SS is typically used over CS. Does anyone have any guidelines? Am I looking at the wrong standard?

 

[monaco8774]

The NACE standard just tells you about the material requirements after you have basically decided whether you are going to use stainless or carbon steel. If its any help we use carbon steel in the discharge lines where there is no chance of condensed water drop out and stainless everwhere you might have condensed water dropping out i.e. downstream of coolers and in the compressor suction lines. This is to keep the general corrosion rates reasonable rather than to combat hydrogen induced cracking then we specify both the carbon steel and stainless lines to NACE. However I do know others who use carbon steel all round. Just as a heads up we find in fields with small amounts of H2S (only 150ppm) and only around 2000 psi discharge pressure that we get quite a problem with fouling from iron sulphide (and iron carbonate if you have CO2) on the compressor impellers, it is worth specifying antifouling coatings on the compressor and bumping up the materials a little to handle potential errosion at the inlets.

 

[Tenoa]

Thank you for the reply, that was at least a step in the right direction. Do you know the justification for those who choose to use carbon steel all round? Is it just water condensate that could cause a potential problem, or could these same problems arise if operating conditions place you inside the two-phase region of the gas envelope? Does the high pressure of the gas play a significant role due to increased metal permeability? Sorry for all the questions, I've been having a hard time getting a straight answer. It seems as though the industry is largely driven by opinions rather than hard fast rules as far as this issue goes. Our customer has not specified their material preference (as most typically do, hence my confusion) and are expecting us to provide a recommendation. Thanks again!

 

[Tenoa]

I have been reading through another post - HIC, SSC, SCC, HCL and HTHA - and I see Nelson Curves are mentioned. Do these apply to H2S? They seem to refer to the partial pressure of hydrogen instead of hydrogen sulfide. We do not have a copy of API 941 in-house, but may be able to porvide justification to purchase it if these curves are really what we need.

 

[SJones]

The governing factor is the presence of liquid water as pointed out by monaco. The amount of CO2 is not stated, but that will also contribute to corrosion in wet conditions. Corrosion models generally stop around the 200 bar mark but you could try running one just to get a feel for the predicted corrosion rate of carbon steel when the gas is wet. Classifying the gas as "wet" is usually defined by an operating temperature within 10 deg C of the water dew point. Incidentally, just "specifying to NACE" won't deal with HIC and SOHIC failure modes. Nelson curves are not what you need - the services of an experienced materials and corrosion engineer would be a better purchase!

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/

 

[monaco8774]

Steve has a good point about getting a good corrosion engineer and I can say that completely unbiassed since I'm not one, its much better to get one before you select the materials than afterwards to explain a failure mechanism.

To answer your question about the justification for using carbon steel all round, those I know who have done it have had mixed results, if the piping is well designed and has no pockets or stagnant areas and their is a proper corrosion inhibition regime in place, carbon steel can be a good inexpensive work horse, however trying not to sound to synical carbon steel is the obvious choice when you are responsible for the design and build costs but not for the operations costs.

 

[Tenoa]

Thank you both for your insights, they were very helpful.

 

[NickelMet]

Tenoa:

Just for your information, the API 941 document (Nelson Curves) is for high temperature hydrogen service. They don't apply to H2S (only) materials considerations or design.

~NiM