Page 263 - Handbook Of Multiphase Flow Assurance

262 10. Research methods in flow assurance

polypeptide. Docking may well be the mechanism of hydrate inhibition by these polymers.
However, there is a contradiction between the difference in performance of these polymers in
inhibiting hydrate formation and similarity of results of the docking study. This suggests that
a more detailed study of adsorption of inhibitors on hydrate in water solution is necessary.

Studying of kinetic inhibitor interaction with water: Solvation of the polymer
in the bulk water

Introduction
Structural changes in liquid water were investigated in the second part of the computer
studies of gas hydrate inhibition. This testing of the second kinetic hydrate inhibition mecha-
nism hypothesis (destruction of structure in the bulk water) was performed using molecular
dynamics. Different polymers or their monomers were solvated in SPC water, depending on
the desired polymer concentration. Effects of each polymer on the structure of water was ob-
tained by comparing it with the structure of pure SPC water at the same conditions.
The work was started with selecting the SPC water model for use after the comparison of
water models available in SYBYIT. Melting point of the SPC water model was fit to 200 K as
reported (Karim et al., 1990) by scaling of the electrostatic interactions. This was followed by
determination of the structure of water at 203 and 220 K (scaled 277 and 298 K). Simulation of
the polymer solutions and analysis of the structure of solvent concluded this work. Initially
the models of water available in SYBYL were verified by comparing the oxygen-oxygen
radial distribution function of simulated and real water at 298 K, with results presented in
Fig. 10.33. The radial distribution function (RDF) measures local density as a function of dis-
tance. RDF can be viewed as the probability of finding a water molecule at a certain distance
from a particular water molecule. Integration of the area under the first peak represents the
number of the nearest neighbors. The best fits of simulation to data were obtained for the SPC
(simple point charge) and the TIP3p (transferable intermolecular potential 3 point) models.
Based on this result the SPC water model was selected. A comparison of oxygen- hydrogen
radial distribution functions for the SPC model and water was made. The oxygen-hydrogen
radial distribution function indicates the average length of hydrogen bonds in water through
the position of the first peak. A good comparison was obtained between the experimental
data (Soper and Phillips, 1986) and the simulation results (Fig. 10.34). This completed the
verification of the water model radial distribution function. The most interesting outcome of
this part of the work is that the kinetic inhibitors affect the structure of hydrogen bonded net-
work of water molecules in such a way as to make hydrate formation more difficult. This was
discovered through the counting of rings in three dimensional hydrogen bonded network of
water in a fashion similar to that described by Rahman and Stillinger (1973). The structure of
water can be described as a network of hydrogen bonds connecting almost all water mole-
cules, with only a small number of molecules free of the network. At higher temperatures, the
energy of hydrogen bonding (approximately 5 kcal/mol) is insufficient to keep the moving
water molecules together and the network of hydrogen bonds becomes very loose. As the
temperature decreases, water molecules vibrate less and a dense network of dynamic hy-
drogen bonds is formed. Hydrogen bonds in this network are arranged in rings which form
and rearrange with time. Rings sizes are measured in terms of the number of participating
water molecules. The most probable sizes of hydrogen bonded rings are 5 and 6. For a com-
parison, water molecules in ice (hexagonal Ih) are arranged exclusively as 6-membered rings,