Hydrate formation in multiphase flow systems presents significant challenges for flow assurance, particularly in crude oil and gas condensate transport. One of the indicators of plugging risk is an increase in slurry viscosity at the onset of hydrate formation. Existing rheology models in literature can predict the steady-state viscosities, however the transient viscosity peaks are not predicted. These peaks observed in rheological experiments can greatly exceed steady-state viscosities and could potentially be the cause of hydrate plugging in the field when sufficient pressure drop is not available to move a fluid with high viscosity. In this study, Camargo and Palermo’s viscosity framework was coupled with a new cohesive force model where interparticle forces are expressed as a function of hydrate and free water fractions derived from experimental gas consumption data. To capture this transient behavior, the model was built with the hypothesis that water-hydrate micro-mechanical forces (order of magnitude stronger than hydrate-hydrate force) are dominant in the initial stages of hydrate volume growth, while gradually shifting to hydrate-hydrate dominant force as free water continues to convert to hydrate. Our approach can capture the underlying physics of particle–particle interactions and the consequential effect on the average cohesive force and fluid viscosity without the requirement of a complex mathematical population balance model. The result is a new framework that treats each point along the viscosity curve as a quasi-steady state, allowing one to track how particle interactions evolve over time. In this study, we have shown that the observed transient rheological behavior can be captured by considering the average cohesive force changing with time. By incorporating parameters such as hydrate-hydrate (measured) and water-hydrate cohesive forces (fitted), the model successfully reproduces both the peak and decay in viscosity with peak viscosity errors lower than 3% and peak timing errors lower than 0.2 h. Incorporation of the trapped water within hydrate particles further improves matching of the experimentally obtained transient viscosity, thereby allowing estimation of the average trapped water from rheology data. This cohesive force-based model provides both a mechanistic insight into hydrate agglomeration and a framework for integration into multiphase flow simulators.
Greener et al. (Thu,) studied this question.