Centrifugal Pumps and Amines 1

Yes. We may many multi-stage pumps pumping lean and rich amine. The main considerations that I would suggest are related to material of construction, mechanical seals and freeze protection. A few of our rich amine pumps were originally purchased in carbon steel. Some of the pump manufacturers may consider this to be adequate. However, ever one of these carbon steel pumps in our plant had to eventually be replaced with stainless steel. Our preferred material is 316 SS (API material class S-8) which holds up quite well in our amine streams.

Mechanical seals in amine can be tricky. Many of our older amine pumps have been converted to dry gas seals. Since you specifically described a rich amine service, the risk of H2S release would dictate the use of a dual mechanical seal. We are not allowed to use API Plan 52 because this requires a vent to flare and our flare system is not suitable for amine in most parts of the plant. An API plan 53 is also a problem. Several of these pumps have suction pressures above 150 psi which makes the use of nitrogen pressure on the seal pot undesirable because nitrogen will go into solution in the barrier fluid and flash out of solution between the seal faces. A dry gas seal avoids these issues as long as you have a dependable source of nitrogen at a high enough pressure to supply the seal. The dry gas seals we have in use have performed quite well.

Our plant is in the northern US and freeze protection is critical. Steam tracing for freeze protection can resulted in hot spots that lead to accelerated corrosion. Electric tracing is very much preferred to protect against freezing without driving corrosion problems.


Johnny Pellin

When sizing centrifugal pumps, do we treat amines as compressible or incompressible?

Thanks for this info above. Its helpful.

Amine would be considered as incompressible. But, there is one other consideration. In our specifications, it specifically states that we cannot consider sub-cooling when calculating NPSH available to an amine pump. In other words, we have to consider the fluid to be at bubble point at the surface of the liquid in the supply vessel. Only elevation head and flow losses are relevant. We don't take any credit for vapor pressure below vessel static pressure. We require this because even though the amine is well below its boiling point, it is saturated with H2S and other gas. From a standpoint of cavitation, the pump will behave as if the liquid is at boiling point in the supply vessel.

Johnny Pellin

Vic123

Most multi-stage manufacture use Monel K-500 shaft material as a standard. This is not a very good choice for amine service. Monel contains 27 to 33 percent copper.

One of the first identified incidents of stress corrosion cracking was termed “season cracking” in brass cartridge cases in jungle environments. Rotting vegetation gave off sufficient ammonia to crack brass cartridge cases but only in the rainy season – thus the term season cracking.

Several Amine applications have used Monel shaft material and got away with it, but corrosion cracking is a strong possibility with K-500.

D23

About the sealing system solution I suggest the plan 53B.

It is suitable in case high suction pressure, since the nitrogen in confined inside the elastomeric bladder of barrier fluid accumulator. Moreover the consumption of Nitrogen is limitated at the volume necessary to charge the bladder.