Glycol is often used for hydrate control in upstream LNG facilities in high volumes (~30-40wt% of water production). Some fraction of this glycol remains in the vapour phase at inlet/shore conditions and hence is carried with the gas through to the Acid Gas Removal Units (AGRUs).  Due to its boiling point being close to that of the aMDEA (the amine used in the AGRU unit) it accumulates in the closed amine loop. This has led to mono-ethylene glycol (MEG) accumulating in the amine in the production trains since start up on some facilities.

The presence of glycol in the amine impacts the ability of the aMDEA to potentially remove CO2 and H2S from the feed gas – eventually leading to breakthrough of acid gases into the gas fed to LNG. At low levels of glycol, this breakthrough can be mitigated by increasing the circulation rate of the amine however, once glycol levels become too high then feed gas rates would have to be reduced with subsequent lost production impact.

Currently, there is no known, or simple/proven method that exists for removal of glycol from amine. The industry approach has typically been to dump the amine inventory and replace with fresh amine (either in one turnaround operation, or over time, otherwise known as a ‘bleed and feed’ approach).  This may be palatable for small amine systems but for larger systems, this can result in the disposal of large volumes that need to be drained and replaced, at significant cost, with supplies often needing to be sourced from Europe.

Reclamation (distillation under vacuum) has been used on smaller amine systems but on larger systems at industrial facilities it has been deemed unmanageable due to the required interfaces with utilities.  It also still produces a waste stream and is not really a ‘set and forget’ process. As a result, such an approach is undesirable because operationally, it is akin to running a whole new process unit.

This CEED project will undertake a ‘deep-dive’ review of the physical properties of the two chemicals and review/investigate potential methods for separation in order to negate the dump & refill program that will be an OPEX burden for any facilities that are operated this way until end of field life. Ideally, the separation process would have limited connections/interfaces to the operating plant (i.e. somewhat of an offline system for purifying the amine, either batch wise, or as a slipstream from the spare amine tank), although there may be other feasible solutions warranting investigation.

This project is to be executed as a ¾ CEED project.