130
5 Nucleation of Gas Hydrates
in the system [47]. Sowa and Maeda reported statistically significant variations in
the level of memory preservation that depended on the sample cells used [65].
Buchanan et al. suggested that the memory effect reported in the literature was in
systems which did not attain full equilibrium after dissociation [66]. The later neutron
diffraction studies by the same authors suggested that the memory effect could be
due to a greater local water density that was influenced by the presence of clathrate
hydrates [67]. More recently, He et al. reported accelerated formation of methane
hydrate and CO 2 hydrate in “used water” when the water was pressurized with the
respective guest gases from 1 atm and suggested that the residual structure present
in water as the reason for the memory effect [68]. Wu and Zhang proposed that the
residual structure in liquid water after clathrate hydrate has dissociated had provided
a mass transfer barrier between water and the guest molecules that led to the memory
effect [69]. Sefidroodi et al. examined the formation of cyclopentane hydrates from
dissociated water over a range of superheating temperatures of up to 5 K and the
heating time of up to 24 h [70]. They reported that a transfer of a small amount of
dissociated water to fresh water induced a memory effect that was comparable to one
in a 100% dissociated water [70]. A residual clathrate hydrate structure was proposed
as a possible mechanism of the memory effect from this observation [70].
The intrinsic stochasticity that is inherent to nucleation was always superimposed
to the systematic difference in the induction times that had arisen from the memory
effect. This limitation has posed considerable difficulty in affirming the existence
of the memory effect or quantifying its magnitude. Bylov and Rasmussen reported
that the induction times of methane hydrate and natural gas hydrate did not become
shorter when formed from the dissociated water, while noting that the limited amount
of their data made it impossible to draw a definitive conclusion [71]. Fandino and
Ruffine studied the formation of methane hydrate at 10 and 19.5 MPa and suggested
that the stochastic nature of clathrate hydrate nucleation overcame the memory effect
[72]. They found no evidence of any statistically significant decrease in the induction times at 7 K of superheating for 5 h [72]. Sefidroodi et al. suggested that the
smallest of the superheating temperatures of 6.5 K used by Fandino and Ruffine
might already have been high enough to destroy the memory effect [70]. Based on
unprecedented numbers at the time of several thousand nucleation events, Sowa and
Maeda concluded that the presence of the memory effect was undeniable (statistically
beyond doubt) [65].
Wilson and Haymet studied the nucleation of tetrahydrofuran (THF) hydrate at
superheating temperatures of up to 15 K using an Automated Lag Time Apparatus
(ALTA) and round no evidence for a presence of statistically significant memory
effect [73]. However, unlike typical guest gases of clathrate hydrates, THF is miscible
with water at all proportions, so any mass-transfer limited issues that are typical
for clathrate hydrates would not arise. For example, supersaturation of THF in a
dissociated water cannot arise.
5 Nucleation of Gas Hydrates
in the system [47]. Sowa and Maeda reported statistically significant variations in
the level of memory preservation that depended on the sample cells used [65].
Buchanan et al. suggested that the memory effect reported in the literature was in
systems which did not attain full equilibrium after dissociation [66]. The later neutron
diffraction studies by the same authors suggested that the memory effect could be
due to a greater local water density that was influenced by the presence of clathrate
hydrates [67]. More recently, He et al. reported accelerated formation of methane
hydrate and CO 2 hydrate in “used water” when the water was pressurized with the
respective guest gases from 1 atm and suggested that the residual structure present
in water as the reason for the memory effect [68]. Wu and Zhang proposed that the
residual structure in liquid water after clathrate hydrate has dissociated had provided
a mass transfer barrier between water and the guest molecules that led to the memory
effect [69]. Sefidroodi et al. examined the formation of cyclopentane hydrates from
dissociated water over a range of superheating temperatures of up to 5 K and the
heating time of up to 24 h [70]. They reported that a transfer of a small amount of
dissociated water to fresh water induced a memory effect that was comparable to one
in a 100% dissociated water [70]. A residual clathrate hydrate structure was proposed
as a possible mechanism of the memory effect from this observation [70].
The intrinsic stochasticity that is inherent to nucleation was always superimposed
to the systematic difference in the induction times that had arisen from the memory
effect. This limitation has posed considerable difficulty in affirming the existence
of the memory effect or quantifying its magnitude. Bylov and Rasmussen reported
that the induction times of methane hydrate and natural gas hydrate did not become
shorter when formed from the dissociated water, while noting that the limited amount
of their data made it impossible to draw a definitive conclusion [71]. Fandino and
Ruffine studied the formation of methane hydrate at 10 and 19.5 MPa and suggested
that the stochastic nature of clathrate hydrate nucleation overcame the memory effect
[72]. They found no evidence of any statistically significant decrease in the induction times at 7 K of superheating for 5 h [72]. Sefidroodi et al. suggested that the
smallest of the superheating temperatures of 6.5 K used by Fandino and Ruffine
might already have been high enough to destroy the memory effect [70]. Based on
unprecedented numbers at the time of several thousand nucleation events, Sowa and
Maeda concluded that the presence of the memory effect was undeniable (statistically
beyond doubt) [65].
Wilson and Haymet studied the nucleation of tetrahydrofuran (THF) hydrate at
superheating temperatures of up to 15 K using an Automated Lag Time Apparatus
(ALTA) and round no evidence for a presence of statistically significant memory
effect [73]. However, unlike typical guest gases of clathrate hydrates, THF is miscible
with water at all proportions, so any mass-transfer limited issues that are typical
for clathrate hydrates would not arise. For example, supersaturation of THF in a
dissociated water cannot arise.