The complex modeling accuracy of gas hydrate models has been recently improved owing to the existence of data for machine learning tools. In this review, we discuss most of the machine learning tools used in various hydrate-related areas such as phase behavior predictions, hydrate kinetics, CO2 capture, and gas hydrate natural distribution and saturation. The performance comparison between machine learning and conventional gas hydrate models is also discussed in detail. This review shows that machine learning methods have improved hydrate phase property predictions and could be adopted in current and new gas hydrate simulation software for better and more accurate results.
In this study, series of non-ionic surfactants from Span and Tween are evaluated for their ability to affect the viscosity profile of cyclopentane hydrate slurry. The surfactants; Span 20, Span 40, Span 80, Tween 20, Tween 40 and Tween 80 were selected and tested to provide different hydrophilic-hydrophobic balance values and allow evaluation their solubility impact on hydrate formation and growth time. The study was performed by using a HAAKE ViscotesterTM 500 at 2 °C and a surfactant concentration ranging from 0.1 wt%-1 wt%. The solubility characteristic of the non-ionic surfactants changed the hydrate slurry in different ways with surfactants type and varying concentration. The rheological measurement suggested that oil-soluble Span surfactants was generally inhibitive to hydrate formation by extending the hydrate induction time. However, an opposite effect was observed for the Tween surfactants. On the other hand, both Span and Tween demonstrated promoting effect to accelerate hydrate growth time of cyclopentane hydrate formation. The average hydrate crystallization growth time of the blank sample was reduced by 86% and 68% by Tween and Span surfactants at 1 wt%, respectively. The findings in this study are useful to understand the rheological behavior of surfactants in hydrate slurry.
CO2 storage as hydrates in porous media is a promising method for storing carbon dioxide (CO2). However, the sluggish formation kinetics of hydrates urge the need to focus on the use of additives (promoters) to accelerate hydrate kinetics. This study investigates the effect of amino acid solutions in brine on CO2 hydrate formation and dissociation kinetics in quartz sand particles QS-2 (0.6-0.8 mm) with 38% porosity. The amino acids l-methionine (l-meth), l-isoleucine (l-iso), and l-threonine (l-threo) were studied at 0.2 wt% using an autoclave hydrate reactor at 4 MPa and 274.15 K in the presence and absence of salt (3.3 wt% NaCl) in 100% water saturation. The hydrate dissociation kinetics was studied at a temperature of 277.15 K. These conditions represent the normal seabed temperature range in Malaysia and hence were used for testing CO2 hydrate formation and dissociation kinetics in quartz sand in this study. Further, CO2 hydrate formation and dissociation experiments were conducted with sodium dodecyl sulphate (SDS) and brine systems as standards for comparison. The findings reveal the best kinetics for l-meth exhibiting the highest CO2 hydrate storage capacity. l-meth recorded a gas-to-hydrate conversion ratio of about 93% at 0.2 wt% in quartz sand with brine. Moreover, l-meth exhibited the lowest hydrate dissociation rate compared to l-iso and l-threo systems, thereby enhancing CO2 hydrate stability in quartz sand. Comparatively, l-meth enhanced the storage capacity by 36% and reduced the induction time by more than 50% compared to conventional promoter SDS in quartz sand with brine, suggesting it to be favorable for CO2 storage applications. CO2 hydrate nucleation time was predicted in quartz sand with and without the best-studied amino acid l-meth system with high prediction accuracy and an absolute average deviation of 2.4 hours. The findings of this study substantiate the influence of amino acids in promoting the storage capacity of CO2 in sediments as hydrates.