Natural gas hydrate formation is a costly and challenging problem for the oil and gas industry. Prediction of hydrates have been carried out through rigorous and laborious solving of mathematical equations called equations of state (EOS) which give accurate results but require appropriate setup and time. Few examples of such equations of state currently used by industry benchmarked software tools include Peng-Robinson (PR), Cubic-Plus-Association (CPA), Soave-Redlich-Kwong (SRK) etc. which more or less provide us with an accurate hydrate stability curve i.e. a pressure-temperature profile for a given composition, which allows us to keep the pressures and temperatures (operating conditions) out of the hydrate stability zone.

Hydrate stability curves are a function of the composition of the fluid (gas) being produced. Compositional changes in the percentage of C1 to C7+ components of gas, would not only affect the specific gravity, but would also change the hydrate stability curve of the gas significantly.

Previous studies have been aimed at finding a quick and precise prediction method for hydrate formation, so as to make swift arrangements to counter any chance of flow assurance issue. Different empirical correlations have been developed on the basis of the composition of the gas being produced that take into consideration the pressure and predict the temperature of hydrate formation. Multiple data points, i.e. fluid compositions from different areas/fields are considered and correlations have been developed to fit the hydrate stability zones of these data points which were found through a more accurate equation of state. As the initial data sets for each correlation are different, the possibility of any two correlations giving the correct and same prediction is very low.

This paper gives an insight into how different empirical correlations like Hammershmidt, Motiee, Makogon, Towler and Mokhtab etc., that have already been derived can be used with better accuracy for a set of different fluid compositions and specific gravities. A sensitivity analysis is done on the performance of each correlation against the accurate hydrate curves found out through the software tool, using different available equations of state. The data points picked here are random and were not included in any data sets adopted for derivation of the correlation.

Furthermore, the mimicked hydrate curve from this new method is cast against the software simulated hydrate curve for a flow assurance steady state simulation study with two deepwater gas wells with different gas compositions. The results of the study suggest that the use of the imitated hydrate curve through analytical approach works well in predicting the hydrate stability zone. It would also not require any software proficiency, would give quick results and would cost a fraction compared to the state of the art simulators.

You can access this article if you purchase or spend a download.