Experiments with gases and liquids, in both static and dynamic conditions, and in the presence of methanol and kinetic inhibitors illustrate the complexity of controlling hydrates in petroleum operations.
This concluding article of a two-part series presents selected results from laboratory experiments with pure and natural gases mixed with distilled water, salt water, and freshwater or salt water containing thermodynamic or kinetic inhibitors.
The first article (OGJ, Feb. 5, 2001, p. 47) described some aspects of hydrates and equations, especially their formation and dissociation.
Experimental procedure
The equipment for the experiments was designed and constructed at Texas A&M University. Hydrate formation was measured under both static and dynamic (active stirring) conditions and the experiments were controlled both visually and electronically.
Included in the experiments were the following gases and liquids:
- Gases-Methane (99.99% ultrahigh purity), carbon dioxide (99.9% CO2 + 0.04% CH4 + 0.06% N2), and natural gas (87.2% CH4, 7.6% C2H6, 3.1% C3H8, 0.508% i-C4H10)
- Liquids-Distilled water, salt water (standard laboratory mixture containing 41.953 g/l. sea-salt composition and meeting American Materials Standard D-1141-52 Formula A, 1.025 density at 15° C.). Table 1 lists the sea-salt composition used to prepare the brine fluid.
- Additives-Thermodynamic inhibitors (2%; 5%; 10%; and 15% methanol), MEG (mono-ethylene glycol), kinetic inhibitors (0.5% or 1%).
For the experiments, cells were either 180 cc with two 64-mm diameter sapphire windows and 500-bar operating pressure or 900 cc with six 92-mm diameter sapphire windows and 210 bar operating pressure. The windows allowed for the observation of hydrate crystal morphology in any location inside the cells.
Many experiments included simultaneous runs in two 900-cc cells, with one cell under dynamic conditions while the other cell was at static conditions. A magnetic stirrer with an external magnetic field and a programmable speed of up to 660 rpm provided the fluid dynamics.
Omega transducers, 0.05-bar accuracy, and thermocouples, 0.05° C. accuracy, measured the pressure and temperature and a computer data-acquisition system, video, and photographs recorded most parameters.
Parameters controlled during the test included:
- Compositions of the hydrate forming gas and water.
- Pressure.
- Temperature and rate of temperature change.
- Degree of fluid supercooling at the moment of hydrate-formation onset.
- Rate of hydrate decomposition as the temperature is increased and pressure decreased.
- Degree of temperature increase at the onset of hydrate formation.
- Pressure drop during hydrate formation.
- Content of dissolved gas in water or an aqueous solution, both before and after hydrate formation.
Measured parameters included rate of hydrate nucleation, rate of radial growth of gas hydrate on the free gas-water surface, thickness of the hydrate film during the hydrate formation at the gas-liquid interface, rate at which the thickness changed with time, hydrate volume, linear growth rate and gas hydrate crystal morphology, mass and volume rates of hydrate growth, and porosity of the massive hydrate crystals.
Also monitored were the intensity of recrystallization, hydrate solubility, and stability of different morphological forms of hydrate during a temperature increase and a pressure drop in presence of thermodynamic and kinetic inhibitors.
Pure methane with distilled water
Fig. 1a presents typical pressure and temperature changes during methane hydrate formation and decomposition in a closed volume using distilled water. Sections of the Curve ABCDEFGHI reflect various stages of phase transitions as follows:
- Line AB represents saturation of water with gas.
- BCD occurs during the cooling of the system without phase transitions.
- CD illustrates supercooling the system with hydrate formation beginning at Point D.
- DEF shows the process of hydrate formation.
- FG shows the data when the temperature increases without hydrate decomposition.
- GH illustrates hydrate decomposition after a temperature increase.
The hydrate starts to decompose at Point G, and ends in H, which corresponds to the equilibrium conditions MN. Line HI shows pressure increases because of increasing temperature and change in gas solubility in water (without phase transitions).
Line MN is commonly used to calculate the amount of inhibitor necessary to prevent hydrate formation and various mathematical models can determine Curve MN. These computations are based on gas and water compositions, pressure, and temperature. We emphasize, however, that MN corresponds to decomposition conditions and not to hydrate formation.
Initiation of gas-hydrate formation usually requires supercooling. The amount of supercooling depends on such factors as gas and water composition, structural state of liquid water, pressure, temperature, surface tension at the gas-water interface, volume of gas dissolved in the water, and amount of water vapor in the gas, or vapor pressure.
In Fig. 1a, greater than 4° C. of supercooling was required before gas hydrate crystals began to form during dynamic conditions. Supercooling reached 6-12° C. for primary hydrate formation at static conditions with freshly distilled water.
Fig. 1b shows gas-hydrate pressure, temperature, and volume that corresponds with data from Fig. 1a. One can clearly see that gas hydrates began forming at Point D and began melting at G.
We emphasize that existing mathematical models1 for determining hydrate-formation conditions or inhibitor concentration for preventing hydrate formation should only be used for finding hydrate-decomposition conditions. These models should not be used to predict the point at which hydrates begin to form.
Natural gas with distilled water
Figs. 2a and 2b show natural gas hydrates under static conditions and the change in hydrate volume with time. Line MN is the equilibrium curve of hydrate decomposition.
To start hydrate formation (Line CD) required a 12.2° C. supercooling. Line DE reflects conditions of hydrate accumulation, and EF shows pressure increase during a temperature increase without hydrate decomposition.
Hydrate started to decompose at thermodynamic conditions at Point F, and decomposed completely at G.
FG represents the equilibrium conditions of dissociation of a hydrate with variable composition during a temperature increase. During that time, the molar composition of the hydrate became heavier, and its mechanical strength decreased.
Fig. 1a and Fig. 2a differ because of the gas composition. Fig. 1a is for a pure methane-distilled water mixture, while Fig. 2a is for a natural gas-distilled water mixture.
As Fig. 1a shows, the hydrate decomposition follows the equilibrium (MN) for the methane-water mixture. But in natural gas-distilled water test (Fig. 2a), the gas hydrates begin to decompose at Point F, at a temperature much lower than predicted by MN.
Again, to accurately use available models such models must be empirically fitted to laboratory measured data.
Natural gas with salt water
Fig. 2c shows hydrate behavior for natural gas in a saltwater mixture under dynamic conditions, while Fig. 2d shows hydrate volume change with time.
Hydrate accumulation and dissociation rates, and crystal morphology all depend on hydrate-former composition, pressure, and temperature.
Under static conditions, Structure I hydrate is more dense and is characterized by a lower diffusive permeability. This hydrate also accumulates under static conditions, but at a significantly lower rate than under dynamic conditions.
In Figs. 2c and 2d, notice that for the natural gas-saltwater mixture, hydrates begin dissociating at Point F, which is at a much lower temperature than the equilibrium line (MN).
Static, dynamic tests
Most experiments were run more than once to ensure reproducible results. Typically, each experiment used two cells: one cell under static conditions and the other under dynamic conditions. Pressure, temperature, and composition were kept the same in both cells.
Fig. 1c shows hydrate volume formed in the pure methane-distilled water test. The dynamic curve in Fig. 1c is the same as the volume curve in Fig. 1b. The static curve in Fig. 1c, however, shows that hydrate volume formed under static conditions is much less than the volume formed under dynamic conditions.
This difference is easy to explain. Under static conditions, a hydrate film forms at the gas-water interface. Once the film forms, additional hydrates can form only as water, and gas diffuses through the solid hydrate film. The hydrate volume formed, therefore, is influenced by the slow rate of diffusion through the film.
Under dynamic conditions, however, the gas and water mix freely, thus allowing more gas hydrate to form. The formation rate depends on heat transfer.
Methanol added
Figs. 3a and 3b show a static test with pure methane, distilled water, and 2% methanol. In this test, hydrate formation required significant supercooling.
The gas hydrates also did not completely dissolve until the temperature was increased well above the equilibrium line (MN).
Line EFG shows that hydrates can form with increasing temperature dependence on active diffusion of water molecules through hydrate film.
Crystal morphology
We identified three hydrate morphological forms: massive, whiskery, and gel-like. The morphology is affected by hydrate-former composition, pressure, temperature, and supercooling.
Massive crystals grow when water and gas molecules sequentially adsorb on a growing crystal. The crystal surface is constantly renewed during this process.
Whiskery crystals grow by adsorption of gas and water molecules moving through pulsating capillary channels between a metal surface and the base of a growing hydrate crystal.
A high degree of perfection, absence of dislocations, low porosity, low diffusive permeability, and high stability characterize whiskery crystals.
Inhibitors
Thermodynamic or kinetic inhibitors influence the formation and decomposition of gas hydrates.
Thermodynamic inhibitors (alcohols, glycols, and electrolytes) lower the chemical potential of water and the hydrogen bond energy, which requires additional cooling before hydrates will begin to form. The inhibitors also reduce hydrate stability.
Kinetic inhibitors and anti-agglomerates do not lower the onset temperature of hydrate formation, but they adsorb on the surface of hydrate microcrystals and significantly alter surface tension at the interface between the hydrate forming phases. These inhibitors prevent a further increase in crystal size and retard formation of large hydrate agglomerates and solid plugs.
By adsorbing on the crystal surface, some kinetic inhibitors cause hydrates to be more solid, thus, increasing the equilibrium temperature of hydrate decomposition.
On the other hand, the presence of any inhibitor decreases the hydrate strength while increasing hydrate porosity, diffusive permeability, and rate of hydrate growth.
One set of tests observed gas hydrate kinetics and morphology with different inhibitors under static or dynamic conditions and at pressures up to 360 bar and temperatures as low as -10° C.
In the presence of methanol, interesting observations made were that supercooling before the onset of hydrate formation is significantly lower, and the rate of hydrate formation is higher than for pure water without methanol.
When methanol is present in water, the formed crystals actively transform from one morphological shape into another as a function of time.
Various hydrate formation inhibitors produce different effects during hydrate formation and decomposition. Thermodynamic inhibitors usually lower hydrate decomposition temperature while kinetic inhibitors usually increase hydrate decomposition temperature.
Kinetic inhibitor, dynamic
Figs. 4a and 4b are for a natural gas in salt water with a 0.5% kinetic inhibitor under dynamic conditions. Lines ABC and DE show the thermal effect on the pressure of gas without phase transitions. Line EFG shows phase transition (hydrate dissociation) with temperature. Line BC shows that an 11.6° C. supercooling was needed to start hydrate formation.
Hydrate formed along Line CD and started melting at Point E, when temperature was increased.
EFG shows that the decomposition of a hydrate with variable composition did not end at Point F, which is the intersection with the equilibrium curve of hydrate stability without an inhibitor. The hydrate did not completely melt until Point G, at 6.6° C. higher than the equilibrium temperature.
The kinetic inhibitor, which was in solution, adsorbed on hydrate crystals, stabilized the hydrate crystals, strengthened them, and decreased the water vapor pressure above the hydrate. As a result, temperature for complete hydrate decomposition increased.
Fig. 4b illustrates a change in hydrate volume over time at the conditions shown in Fig. 4a.
Kinetic inhibitor, static
Figs. 4c and 4d illustrate hydrate behavior for natural gas in salt water with 0.5% kinetic inhibitor under static conditions. Lines ABC and DE show the thermal effect on the pressure of gas without phase transitions.
Line BC shows the supercooling before the start of hydrate formation. At Point C hydrate begins to form. It starts to melt at Point E, when the temperature was increased.
Line EFG shows the hydrate decomposition as the temperature increased. With a kinetic inhibitor, the hydrate decomposed completely at 8.2° C. higher than the equilibrium temperature for that pressure.
Again, the hydrate lattice strengthened because of the kinetic inhibitor adsorbed on the hydrate surface. The rate of temperature increase during these tests did not exceed 1° C./day.
Fig. 4d shows the hydrate volume for this test.
In both static and dynamic tests with a kinetic inhibitor, hydrate formed only in the gas phase through evaporation and subsequent condensation of water vapor. Hydrates formed very slowly in the aqueouse phase.
Furthermore, an important observation was that the hydrate volume formed under static conditions was greater than the volume under dynamic conditions.
This observation contrasts with the conditions shown in Fig. 1c. Under those conditions, the hydrate volume formed was much greater under dynamic conditions because mixing constantly created new surfaces between the gas and liquid. With kinetic inhibitors, however, more hydrates formed under static conditions.
In tests under dynamic conditions, when inhibitors were added to the liquid phase, a small hydrate volume formed quickly as microcrystals. The inhibitor quickly adsorbed to the surface of the microcrystals and prevented further growth. Thus, hydrate volumes were limited.
Under static conditions, these same microcrystals formed a film at the gas-liquid interface. The film was very weak and very permeable. As such, water molecules without inhibitor easily diffused through the film and into the gas phase, in which hydrate crystals grew rapidly.
Factors affecting hydrate formation in the presence of kinetic inhibitors include vapor pressure, surface tension, and active sorption, which must be evaluated in more detail to explain our experimental results. As with many research endeavors, some experiments offer knowledge that leads to many more experiments.
Other kinetic inhibitor
Fig. 5 is for natural gas in salt water with another class of kinetic inhibitor that stabilizes the already formed crystals by elevating the decomposition temperature of the entire hydrate mass. In these figures, Lines ABC and FGH show a thermal effect on gas pressure without phase transitions on the static and dynamic systems.
Line BC shows supercooling before the onset of hydrate formation. Hydrates started to form at C and continued along CDEF. GH shows the superheating required for hydrate decomposition in both systems, and MN are the equilibrium lines of hydrate dissociation.
Figs 5a and 5c show that hydrate formation required a 7.8° C. supercooling under dynamic condition and a 14° C. supercooling under static conditions. Figs. 5b and 5d indicate the hydrate accumulation and melting rates during these cycles. These data show that with these kinetic inhibitors hydrate forms much faster and melts at less superheating under static conditions than under dynamic conditions. Also, the hydrates formed under static conditions are softer and less dense than the hydrates formed under dynamic conditions.
The tests found two hydrate metastability zones: a zone of hydrate formation metastability and a zone of hydrate decomposition metastability. The curves in Fig. 6 characterize the metastability zones for hydrate formation conditions (left of hydrate equilibrium), and for hydrate decomposition conditions (right of the equilibrium curve).
As stated earlier, the supercooling required to start hydrate formation was as much as 12.5° C., depending on the state of the aqueous phase, the gas composition, the rate of cooling, fluid dynamics regime, and other parameters.
The metastable zone of hydrate decomposition in the presence of an inhibitor was determined by the inhibitor's activity, the magnitude of the water vapor pressure decrease, and the lowering of the water chemical potential in the presence of inhibitor adsorbed on the hydrate crystal surface.
Reference
- Van der Waals, J.H., and Platteeuw, J.C., "Clathrate Solutions," Advanced Chemistry Physics, Vol. 2 (1959), No. 1.