Characterizing Regenerated Mono-Ethylene Glycol for Methane Hydrate Management

Abstract

Gas hydrate formation and deposition are of critical concern for high-pressure natural gas production lines. Complete prevention of hydrate formation using high dosages of thermodynamic hydrate inhibitors such as mono-ethylene glycol (MEG) is typically undertaken: this is a multimillion dollar annual cost for each asset. An under-inhibition strategy utilizing MEG at dosages below the full thermodynamic inhibition requirement offers a cost-effective alternative, achieving a transportable hydrate slurry while maintaining safe operations. There is a limited understanding of under-inhibited systems and a lack of simulation tools to reliably predict hydrate blockages, which have been major barriers to deploying hydrate management strategies. In the current work, we investigate the impact of regenerated MEG samples from a live plant with a 30 year operating history on the thermodynamics, interfacial characteristics, kinetics, and transportability of hydrate formation. A high-pressure micro-differential scanning calorimeter (HPμ-DSC), an optical interfacial tensiometer, and a high-pressure sapphire visual autoclave (HPVA) were employed. HPμ-DSC results indicated that regenerated MEG retained its inhibition efficacy. Further, results obtained from the tensiometer showed that this MEG sample is contaminated with surface-active species. These possess an adsorption affinity for the oil–water interface: the oil–water interfacial tension decreased by 40% compared to a paraffin oil baseline with only 0.1 wt % regenerated MEG, and up to ∼90% reduction was observed with 5.0 wt % regenerated MEG. HPVA tests showed that the addition of regenerated MEG accelerated the initial rate of hydrate formation but resulted in a 50% reduction in the torque, indicating an increased hydrate transportability. Regenerated MEG at a mass fraction of ≥10 wt % in the aqueous phase generated a transportable hydrate slurry at operating conditions of 8.0 MPa pressure and 274.2 K. This implies an approximately 65% reduction below the MEG content required for complete inhibition.