Centrifugal compressor with 95% H2S

[Nicolas13]

Would you use an axially or radially CC or integrally geared compressor containing 95% H2S and light hydrocarbons? Pressures barg in/out: 2 / 12. Power less than 1Mw. Dry gas seals. What would be your basic requirements?

 

[rotw]

Are you sure about the 95% H2S and is this figure % mol. weight?
To know the severity you need to know the pressure but also water content. You need also to know if you have CO2 and in which percent as you could have a risk of sulfide cracking cracking combined with corrosion (acid gas).

Your first check should be against NACE std. and sour service material severity diagram.

If its dry than its dry but its rarely 100% dry unless you have a dehydration unit just before the compressor unit and even thought the unit is subject to failure. So you need to design for worst case scenario.

Typically an axially split machine would be limited to quite lower H2S percentage (we talk here "ppm").
I would not select an axially split nor an integrally geared (due to reliability concern).

I would provide in my specification an accurate gas composition (as much as possible), settle out conditions and explicit all operating modes to the compressor manufacturer and let them propose what is most adequate as design. The main advantage of an axially split versus radially split is the maintenance convenience. An equivalent axially split machine would probably need to be designed for higher maximum allowable working pressure for same gas and operating conditions so this would increase potentially your costs. Unless there is a real constraint on maintenance for barrel type (inner bundle removal), or a specific requirement from the licensor, if licensed design - barrel type compressor will keep you on safe side. I skip here the implications on the inner parts like impellers etc.

Amongst other things, you would need to specify the testing requirement with respect to the H2S service and conditions. The DGS need also to be designed in accordance. You may need to check what can be allowed to leak to the flare/atmosphere vs. codes / standards as well as local regulations. There are various applications where double type dry gas seals need to be selected in order to prevent any leakage to atmosphere.

In short you need to provide a very exhaustive input to compressor manufacturer in order to let them design a suitable solution. That would be the basic requirement. Last but not least ask for references...as a risk mitigation (this is minimal).

 

[Nicolas13]

Gas does not contain water, I agree to consider this possibility, CO2 less than 0.1%, H2 less than 1%, all concentrations are % weight.

 

[rotw]

You seem to confirm the 95% H2S implicitly. Thus your H2S partial pressure can be in excess of 10 bar. Right?
Funny that you ask for basic requirement...because I am telling you: I don't think you can find on earth a compressor that would compress your gas using any type of conventional technology available in the market.

 

[georgeverghese]

From a process engineer's perpective, one would ask why you would want to go axial - for a given head, flow range is limited and hence there is little possibility of power / utility savings on turdown. Basically speed turndown to match varying process feed flows at a given fixed head is more constrained than in a regular radial feed type centrifugal.

 

[rotw]

The OP is not requesting to go for axial flow, they refer to the split plan (construction) of the centrifugal compressor whether it is vertical (radial) or axial.
As general note, regarding your comment switching to axial could increase the efficiency by few points versus centrifugal machine when operating at high design flow coefficients. Few points of efficiency can mean significant power savings and this can compensate operating range reduction especially when your operating points are not wide spread across the map. Other note: I am not aware of widespread usage of axial flow application with high H2S content.

 

[georgeverghese]

Rotw, Your response has shown me the error in my reading of the OP's post now.
At least we agree on the turndown capabilities of an axial compressor.

 

[Nicolas13]

This gas is already compressed by several reciprocating compressors. For reliability reasons plus the increase of capacity the question of new compressor type is raised.
About the axial type, the suction volume (7000Nm3/h) is too low.
Thanks for comments.

 

[rotw]

My comments relate to centrifugal.
Recips is different story. Just in case it was not clear.

By the way, what is the application / plant where this is to be installed ?

 

[Nicolas13]

Plant is in Europe

 

[rotw]

Thanks for the valuable info.

 

[georgeverghese]

With 95% H2S, obviously one would go for both fuel gas purge on the inboard side and buffer inert gas purge on the outboard side for all shaft seals.