Page 295 - Handbook Of Multiphase Flow Assurance

294 10. Research methods in flow assurance

However, this shift prevents hydrate lattice from collapsing in terms of kinetic stability.
The cause of hydrate instability was attributed to the increase in low and intermediate fre-
quency oscillatory motions of hosts. In their subsequent study (Tanaka and Kyohara, 1993b)
compared intermolecular host-host vibrations in hydrates of propane, xenon and methane. It
was shown that motion of propane guest in a large hydrate cavity does not couple with the
host motion. Thus the fixed lattice approximation can be used to simulate methane and xenon
guests motions inside large cages.
Effect of pressure on the stability of xenon and argon hydrates was reported by Tanaka and
Nakanishi (1994). Chemical potential of water was calculated from the general partition func-
tion and used in the generalized van der Waals and Platteeuw model to calculate hydrates
stability. At high temperatures good agreement with experiment was found.
In a simulation of hydrate stability encaging highly polar guests (Koga et al., 1994) it was
found that amine hydrates are stable because guests don't form hydrogen bonds to hosts,
while in alcohol hydrate many guest-host hydrogen bonds are formed. In an MD simulation
using the SPC water model Westacott and Rodger (1994) showed that the stability of sI hy-
drate calculated in terms of Gibbs free energy of the host lattice increased with the number of
guests. Shape of the hydrate lattice was not affected by the inclusion of guests in small cavi-
ties, but changed when the large cavities were filled with Lennard-Jones guests. Guest-guest
and guest-host radial distribution function (RDF) and cavity RDF (CRDF) were used to show
that guest-host repulsions are more important than the attractive interactions for use with
van der Waals and Platteeuw theory (Rodger, 1994a,b).
Wallqvist (1994) studied the effects of methanol on sI hydrate unit cell stability using MD
with SPC water model. From calculated CRDF he showed that guest molecules increased
hydrate stability, compared to empty lattice. Addition of methanol molecules to the mixture
broadened and lowered the CRDF which indicated a decrease in hydrate stability.
Thermodynamic stability of ethane, ethylene, and CO 2 hydrates (Kvamme and Tanaka,
1995) was found to be in qualitative agreement with experiments in the sub-zero tempera-
ture region. The addition of a virial coefficient to the equation for calculation of guest chem-
ical potential matched the predicted pressures with experimental values for ice-hydrate
equilibrium region. At water temperatures predicted values were in a qualitative agreement
with data.
A MD simulation of sII hydrates (Hirai et al., 1996a) was performed to investigate stabil-
ity of CO 2 hydrate relative to argon hydrate and an empty hydrate lattice. The authors ob-
served broadening of the cavity RDF and higher MSD for host molecules in hydrate with CO 2
guests, compared to reference systems. Although the host molecules retained their positions
throughout the simulation (no melting has occurred), the conclusion was that CO 2 hydrate
will be thermodynamically unstable. This conclusion appears questionable.

Early modeling of hydrate growth

The dynamic growth studies reviewed in this section are questionable attempts at mod-
eling real-world hydrate growth. Computer modeling of hydrate growth is nonsense due
to the simulation times involved—much too short compared to real world growth times.
Rates of axial volume-diffusion hydrate growth reported by Makogon (1981) range from ca.
0.6 to 3 mm/h for different gases which is 1600–8300 Å/s. Molecular dynamics computer