Page 308 - Handbook Of Multiphase Flow Assurance

Experimental study of hydrate crystal growth 307

FIG. 10.66 An octahedral crystal of THF hydrate formed at 1 °C and then placed in a melt with 0.5 wt% PVP
exhibits a planar continued growth (supercooling of 3.4 K).

a {111} surface has six-fold symmetry, while the crystal itself has a three-fold symmetry axis
normal to that plane.
A second possible mechanism might appeal to a time-dependence of the inhibitor adsorp-
tion. If adsorption requires reorientation of the inhibitor molecules to conform with some
feature of the face, then there may not be time for that to occur at the rapidly growing plate
edges, whereas the inhibited plate faces remain slow-growing. This has some precedent in
the effect of kinetic inhibitors on ice growth. It has been reported that one of these changes
the growth habit completely in free growth at low supercooling but does not change it in the
faster growth at high supercooling (Harrison et al., 1987), and Knight has observed the same
phenomenon for several other of the kinetic ice “antifreezes”. It is hard to relate this effect
to polymer diffusivity in hydrate melt since the diffusivities of kinetic inhibitors in water are
unknown.
Preliminary experiments have shown that the plate-like habit found at low concentrations
of kinetic inhibitors is a transition to complete inhibition of crystal growth at higher inhibitor
concentration. Complete inhibition is also a function of supercooling, and a long series of ex-
periments was performed by Roar Larsen at CSM to quantify this as a function of type, molec-
ular weight, concentration of the inhibitor, and the THF concentration in the water solution.
During these tests a single crystal of THF with an octahedral shape was grown in the aque-
ous THF solution. Then this crystal was transferred into the aqueous THF solution containing
a kinetic inhibitor. As one example, 0.75 wt% VC-713, of 70,000 molecular weight, inhibits
growth completely (growth rate undetectable: less than about 0.01 mm/h over 25 h at a su-
percooling of 1.4 K) where the uninhibited linear growth rate of the octahedral crystals was
about 5 mm/h (Larsen et al., 1996).
As controls for the experiments with inhibitors, we performed growth experiments with
other additives that do not provide kinetic inhibition: polyvinyl alcohol, urea, hydroxyethyl-
cellulose, and polyacrylamide (Long et al., 1994). These chemicals were chosen from a list of
non-inhibitors because they are soluble in water. Two of these also have a vinyl backbone and