Evaluation of Different Hydrate Prediction Software and Impact of Different MEG Products on Gas Hydrate Formation and Inhibition

Abstract

Abstract New hydrate profile correlations for methane gas hydrates were obtained computationally (using three different hydrate prediction software packages) and experimentally (with three different MEG products from different suppliers). Methane gas with pure distilled water was the benchmark case used for the software comparison at pressures of 50 to 300 bar. In order to compare the hydrate inhibition performance of the MEG products, aqueous 10 wt% MEG solutions were tested using the isobaric method at a pressure range of 50 to 200 bar. Furthermore, the kinetics of MEG hydrate inhibition were studied experimentally for methane gas using a stirred cryogenic sapphire cell. Hydrate formation start, hydrate dissociation initiation and hydrate dissociation end points were identified and analysed. The results were correlated with the hydrate formation start points predicted by three well known selected hydrate prediction software packages (which all use the Peng-Robinson equation of state). Moreover, the hydrate inhibition performance of the three MEG products was evaluated to determine the superior MEG product that provides the best hydrate inhibition performance. Our analysis shows that the hydrate formation points predicted computationally are not identical to the hydrate formation start points measured in this work. Software "A" and software "B" predicted results matching with the average curve of the experimental hydrate formation start and hydrate dissociation start points, and with a deviation value of 0.06 °C for software "A" and a deviation value of 0.03 °C for software "B". However, software "C" predicted results almost identical with the experimental dissociation start points, and with an average deviation value of 0.54 °C. The methane gas hydrate profiles for the three different MEG products (X-MEG, Y-MEG and Z-MEG) indicated that X-MEG was the most efficient inhibitor as it shifted the hydrate curve most to the left; X-MEG shifted the hydrate formation curve by an average temperature of 2.07 °C when compared to the benchmark curve (100% water); while Z-MEG shifted the curve by an average temperature of 1.81 °C and Y-MEG shifted the curve by an average temperature of 1.71 °C. We conclude that not all software packages predict the same results although they are all based on the same equation of state. Furthermore not all MEG products supplied have the same hydrate inhibition efficiency. Importantly, choosing the best MEG supplier will reduce the OPEX by reducing the amount of MEG used, and it will accommodate more relaxed operating conditions of lower temperatures and higher pressures.