EG vs. TEG added to refrig plant

[Zoobie]

At a gas plant we have two trains. One train dehydrates the gas with a TEG contactor (9 trays). The other injects EG at the inlet of the refrigeration plant and takes it ou from a 3 phase low temperture separator. The operator has noted that the EG system has a lot higher glycol make-up requirement. Without taking a closer look at it, I presume that this is because the liquid-liquid separation of the glycol/water phase from the hydrocarbons is not as good as the the gas-liquid one. My question: is there any reason to believe that TEG would work better than the EG in this application?

 

[Chemeng1122]

Better a late response than never?

Everything else being equal, EG losses will be higher than TEG losses because of vapour pressure differences.
On a more practical thread though, you are correct in a way by saying that the separation efficiency in the refrige LTS is not as efficient as a vapour liquid separaion in the TEG contactor.

One of the most common reasons for glycol losses in a refrige plant is because of a high rich solution viscosity. This can be cause by a number of reasons, one being over regeneration of the lean glycol (<20% water), contamination (we have seen plants d/s of amine units with >40% amine, which is much more viscous at low temps than glycol), and high injection rates resulting in low rich solution water content. Compunding with high injection rates is a low residence time in the LTS which inhibits effective phase separation.

With poor separation in the LTS, you can lose glycol directly to your LPG product stream, or by carrying hydrocarbons down though to the glycol regen loop. The HC's will flash off in the regen, greatly increasing the vapour flow rate in the tower, carrying glycol out the top. We saw at one plant where they condensed the regen vapours into a tank that there was >50% glycol in the tank, which is supposed to be 100% water (or close to).

Another risk of HC carryunder is fouling of the regen packing. This can result in significant losses as well from flow restrictions in the tower.

Andrew Lechelt
Technical Support Engineer
Quadra Chemicals

http://www.quadrachemicals.com