MDEA problem 2

[imans]

Our refinery has replaced DIPA with MDEA in the ground that MDEA has better performance in term of degradation due to CO2 and of H2S loading. After three mount what we found was far from what literaturs said and other refiner's experiences. With DIPA, we could achieve 10 ppm of H2S at treated off gas and the make up rate was about 2 drums (400 lt) per every two weeks. But with MDEA, the H2S content on treatet off gas is about 40-50 ppm and the make up rate is twice than DIPA. Does anyone have similar experience with us? We had asked the supplier to help solving the problem. It hasnt solved yet up to know.

 

[Samiran]

MDEA should be better than DIPA as MDEA is a tertiary amine and is more effective to remove H2S. Have you reduced your circulation rate ? Has foaming been taking place ?

Samiran

 

[25362]

DIPA is a secondary amine while MDEA is a tertiary amine. The latter reacts more slowly with CO₂ forming bicarbonate. With DIPA, CO₂ forms carbamate. MDEA is also considered more selective towards H₂S than to CO₂ resulting in a better composition of the feed to a sulfur recovery unit. It is true that DIPA degrades severely wiht CO₂ to form 3-(2-hydroxypropyl) 5-methyl oxazolidone, which doesn't have acid gas removal properties.

Although I cannot give an explanation of the reasons of your plant's performance, I can submit herebelow a published comparison between the two amines:

DIPA MDEA

Molecular weight 133.2 119.2
Typ. concentration, mass% 30-32 35-50
Typ. lean loading, mol/mol 0.05 0.01
Typ. rich liquid loading, mol/mol 0.15 0.12
Typ. rich gas loading, mol/mol 0.40 0.45
Typ. steam use lb/gal 1.1 1.0
Pure amine relative price 0.95 1.0
Heat of reaction with CO2, Btu/lb 550 475
Heat of reaction with H2S, Btu/lb 475 455

The amine capacity at 200 lb-mol/h acid gas:

Acid gas pickup, mol/mol 0.35 0.44
Concentration, mass% 32 50
Circulation rate, gpm 476 217
Estimated steam required, lb/h 31,400 13,000

General notes:

1. Upon purifying and reclaiming, carbon absorption and filtration are needed to remove not only contaminants before they react with amine but also solids that may stabilize foam.

2. High contactor temperatures reduce amine effectiveness in absorbing acid gas. These temperatures are determined by: a. the incoming lean solution; b. the incoming gas stream; c. the amine content in the circulating solution;
a high amine content solution contains less mass to distribute the heat of reaction, thereby increasing the contactor temperature. Addition of water to the amine solution (diluting it) cools the contactor and reduces potential corrosivity.

3. On the other hand increasing amine concentration will:

1. reduce circulation rate (less water in the solution)
2. increase corrosivity due to the higher boiling temperature which also increases degradation and utility costs
3. increase hydrocarbons' pickup because due to lower water concentration the affinity to hydrocarbons increases
4. increase solvent vaporization losses
5. increase foaming tendency due to increased organic content.

Thus the benefits of decreased pumping costs should be compared with increased amine losses, foaming and possible corrosion.

Good luck.

 

[imans]

We've tried to reduce the circulation rate but H2S content on the treated off gas increased. All operating condition seemed normal. No foaming was detected.

 

[Samiran]

It appears you've changed the solvent without making any mods to the column. In this case you would have to readjust your operating parameters till you get your desired result. Theoretically ( and also practically) MDEA will require less circulation rate however do incresae that if it helps.

Samiran

 

[bspooner]

Imans - from what I can tell this is a low pressure application where the intent is to selectively remove H2S over CO2. MDEA will do this better than DIPA but due to the low partial pressure of H2S, there is little driving force to get it into the MDEA. You need to focus on two things: keep the MDEA temperature low (30-35 C, if I am guessing your application correctly there should be no hydrocarbons in the gas stream and therefore little chance of foaming) and the lean loading needs to be extremely low. Run your regeneration system with a reflux ratio of 6 or more.

As for losses - to me they seem very low to begin with, and a doubling of already low losses would still be considered acceptable in most systems. There is not a significant difference in vapor pressure between MDEA and DIPA and an increase in inventory loss is most likely either due to foaming (unlikely), or mechanical damage leading to solution carry over.

Ben Spooner, P. Eng
Process Engineer
Amine Experts Inc
Calgary, Alberta

 

[npk]

MDEA is a tertiary amine & hence a better one for selectively stripping H2S compared with the mono & di amines.
some parameters which may help in acheiving better results:
1) Gas temperatures between 30-35'C.
2) Amine temperature of 35-45'C with a
temperature differential between the Gas &
Amine being 5-10'C.
3) Lower amine circulation rates
4) If the Abosorber column is designed for multi-
tray liquid feed, then you can try with the
lowest feed tray first and then you can move
up in steps, to minimise losses of amine due
to carry over.
5) Better regeneration of amine by maintaining
reflux of between 5-8.
all the best

 

[dcasto]

You need to give a little more on operating conditios and CO2 content.
Charcoal filters are a must in MDEA and normally not used in DIPA and if ther they were small.

 

[warbag]

As opposed to the packed tower, I am going to vaporize the MDEA/Water solution and contact the H2S with MDEA at approximate temp. 50 C and extremely low pressure in a 18,000 fpm atomizer, then instantaneously (app. 1/300 of a second) pass the mixture through multiple low (venturi) pressure zones and release it into a pressure vessel for re-entrainment of liquid separating it from remaining air. Will this process defeat the desired capture of H2S into the liquid or enhance it? I hope to be able to use less amine than customary by intimately colliding the molecules without the interference of surface tension. What do you think will happen?

 

[warbag]

Correction: The temp. is supposed to read 25 C instead of 50 C