Synthesis Procedures And Apparatus 21 Pure methane hydrate

Samples of polycrystalline methane hydrate can be efficiently synthesized by promoting the general reaction CH4 (gas) + 6H2O (sol.—>liq.) —> CHi*6H20 (sol.) (Fig. 1). This product is achieved by the mixing and subsequent slow, regulated heating of sieved, granular, H20 "seed" ice and cold, pressurized CH4 gas in an approximately constant-volume reaction vessel (Fig. 2). This seeding method for hydrate nucleation and growth permits successful synthesis of polycrystalline test specimens with controlled and uniform grain size with no detectable preferred crystallographic orientation, which we have verified through sample-replica observation and powder x-ray analysis (Stern et al, 1998a).

Sample fabrication details are as follows: CH4 gas from a source bottle is initially boosted in pressure (P) by a gas intensifier and routed into sample

Figure 1: Methane hydrate synthesis and annealing conditions in relation to the CH4-H20 phase diagram. Shaded region shows methane hydrate stability field. The metastable extension of the H20 melting curve is shown by the solid grey curve. Black dotted lines trace the reaction path during methane hydrate synthesis from ice + gas mixtures. Points A-F indicate the P-T path during reaction. For synthesis, H20 "seed" ice at 250 K (pt. A) is pressurized with CH4 gas to 25 MPa (B). Heating the mixture through the H20 melting point (C) and up to 290 K (C-D-E) promotes full and efficient conversion of the ice to gas hydrate. Samples are then cooled to 250 K (F), and can then be quenched (F]) and removed from the apparatus, or annealed at conditions closer to the equilibrium curve (F2, F3). The "n" numbers represent hydrate stoichiometry under various conditions.

molding vessels housed in a conventional deep freezer. The sample assembly (Fig. 2) consists of a manifold of as many as four stainless steel vessels immersed in an ethyl alcohol bath initially held at freezer temperature of ~ 250 K. One vessel serves as a reservoir to store and chill pressurized CH4 gas, and the others contain the sample molds. Each mold consists of a hollow split-cylinder that encases an indium sleeve filled with a measured mass of H20 "seed" ice typically packed to 40% porosity. Seed material is made from a gasfree and nearly single-crystal block of ice grown from triply distilled H20, crushed, ground, and sieved to 180-250 pm grain size (Durham et al, 1983). The sample vessels with seed ice are initially closed off from the reservoir and evacuated. A disk inserted on top of the packed ice grains prevents displacement of the ice during evacuation. Multiple thermocouples inserted into the base of either sample prior to loading of the ice permit careful monitoring of the sample's thermal history during synthesis and subsequent testing. A Heise pressure gauge and pressure transducers monitor gas pressure on the samples.

The reservoir vessel is first charged with pressurized CH4 gas to 35 MPa and cooled to 250 K. When fabricating a single sample, the reservoir is opened to the pre-evacuated sample chamber and methane pressure (Pcm) drops to ~ 24 MPa. For fabrication of multiple samples, the reservoir charging, cooling, and opening procedures are repeated to bring the larger volume of the multiple-sample system to about 25 MPa at 250 K. These steps serve to fill the porosity between the ice grains to a molar ratio of CH4 to H20 well in excess of that required for complete hydrate reaction. The bath temperature can then be slowly

raised by various methods such as by use of a ring immersion heater or a simple hot plate located beneath the alcohol bath. As the samples and gas reservoir warm, they self-pressurize by thermal expansion. Up to 271 K, methane gas pressure (Pcm) increases approximately linearly with increasing temperature, following a slope governed primarily by the equilibrium thermal expansion of free CH4 in the system. Measurable reaction begins as temperature rises above 271.5 K (the approximate melting point of ice at our synthesis pressure), and consumption of CH4 gas by hydrate formation slows the rate of Pch4 increase (Fig. 3). Progress of the hydrate-forming reaction is monitored by observing the deflection of P from the initially linear P-T curve. Completion of reaction is efficiently achieved by steady heating to ~ 289 ± 1 K over a heating time interval of about 12 to 15 hours after the sample vessel crosses the 271.5 K isotherm (Fig. 3). Data-acquisition software (LabVIEW™, National Instruments) monitors and records the P-T conditions throughout each run. The extent of reaction can be determined mid-run by the measured Pch4 offset from the reversible CH4 expansion curve.

Figure 2: Left: Apparatus for synthesizing test specimens of pure, polycrystalline methane hydrate from CH4 gas and H2O ice by methods described in text. Right: Postulated synthesis model. Numbered arrows (1-4) are as follows: 1) Cold, pressurized methane gas (35 MPa, 250 K) is admitted to the ice in the sample chamber. The reactants equilibrate to -25 MPa and 250 K. 2) Warming above the H2O melting point initiates measurable hydrate formation along the surface of the ice grains, creating composite grains in which a mantle of hydrate envelops an unreacted ice (± melt) core. The reaction rate (diffusion or transport controlled) slows as the hydrate rind grows and thickens, and the inner unreacted core shrinks. 3) Slowly raising the temperature to 290 K promotes further reaction. 4) By the end of the heating cycle, the reaction has reached completion.

Figure 2: Left: Apparatus for synthesizing test specimens of pure, polycrystalline methane hydrate from CH4 gas and H2O ice by methods described in text. Right: Postulated synthesis model. Numbered arrows (1-4) are as follows: 1) Cold, pressurized methane gas (35 MPa, 250 K) is admitted to the ice in the sample chamber. The reactants equilibrate to -25 MPa and 250 K. 2) Warming above the H2O melting point initiates measurable hydrate formation along the surface of the ice grains, creating composite grains in which a mantle of hydrate envelops an unreacted ice (± melt) core. The reaction rate (diffusion or transport controlled) slows as the hydrate rind grows and thickens, and the inner unreacted core shrinks. 3) Slowly raising the temperature to 290 K promotes further reaction. 4) By the end of the heating cycle, the reaction has reached completion.

Following full reaction, the heat source is turned off, and the system slowly cools through the freezing point of ice and down to freezer temperature

(-250 K). At this stage, the P-T record should be inspected for any indication of a freezing anomaly during the cooling cycle, signifying incomplete reaction. Such an anomaly appears as a discontinuous jump in both the P and T readings at several degrees below 273 K, produced when small amounts of unreacted and supercooled water suddenly freeze to form ice (see Fig. 6 in Stern et al., 1998a, for example). The exothermic process of freezing even very small amounts of water (< 1.5 vol. %), coupled with the associated volume increase, is easily measured by the internal thermocouples and pressure transducers. The presence of such discontinuities requires cycling the sample through the melting point of ice a second time to insure completion of reaction.

Figure 3. (A) P-T history of a methane hydrate synthesis run (with two samples) during the reaction: CH4 (g) + 5.9H20 (ice-^melt) CH4*5.9H20. Warming the ice + gas mixture above the H2O solidus (dot-dashed line) initiates measurable reaction. Increasing temperature slowly to 290 K, over an 8-15 hour span, promotes full reaction. Complete reaction is marked by a known pressure drop (APr) from start to finish relative to the extrapolated subsolidus P-T curve. (B) Temperature-time profile during hydrate formation. The thermal profile of the sample as it warms through the H20 melting point is expanded in the inset. Buffering of the sample thermocouples at the H20 melting point indicates that a measurable fraction of melting of the seed ice occurs over this 1.2 hour stage. Full conversion of the H20 to hydrate requires 8-12 hours however, and APr over the early melting stage is not as large as that predicted for complete melting of all residual ice (see Fig. 3A in Stern et al., 1996, for further details).

If no refreezing anomaly appears, samples may then be annealed, tested, or removed from the apparatus. For removal, samples should be vented to low PCH4 within the hydrate field 2 MPa at 250 K), then slowly depressurized while cooled with liquid nitrogen. Pressure must be maintained on the sample to keep the gas hydrate within its stability field during the initial phase of the cooling procedure down to 190 K, but the sample should be fully vented before it cools through 109 K, where any residual pressurized methane in the pore space will liquefy. Allowing liquid methane to settle and freeze as the sample is further cooled to liquid nitrogen temperature (77 K) results in a final material

Nitrogen Gas Hydrate

250 260 270 280 290 Temperature, K

10 15 20 25 Time, hours

250 260 270 280 290 Temperature, K

10 15 20 25 Time, hours

that is mildly explosive and very difficult to work with when rewarmed. Following the quenching procedure, the vessels can be disconnected from the apparatus and opened. The inner, hollow split-cylinders containing the jacketed samples are pushed out of the vessels and pried off the samples. Samples are then wrapped tightly in A1 foil and stored in or directly above liquid nitrogen.

that is mildly explosive and very difficult to work with when rewarmed. Following the quenching procedure, the vessels can be disconnected from the apparatus and opened. The inner, hollow split-cylinders containing the jacketed samples are pushed out of the vessels and pried off the samples. Samples are then wrapped tightly in A1 foil and stored in or directly above liquid nitrogen.

Figure 4. Cylindrical test specimens of polycrystalline methane hydrate and mixed sediment aggregates grown by methods described in the text. Left: Sample of pure methane hydrate (CH4»5.89H20), 2.54 cm x 11.5 cm, with uniform grain size of -300 microns and approximately 28% porosity. The encapsulating indium jacket in which the sample was grown has been split and peeled back from the sample. Right: Pure methane hydrate + sediment aggregates fabricated with pre-specified characteristics. Samples are 2.54 cm by 10 - 11.5 cm. Sample at left is methane hydrate + pure quartz sand, varying the proportion of sand in each quarter of the sample. Middle and right samples are mixtures of pure methane hydrate and 1.25 ± 0.25 mm black AI2O3 particulate. The two samples containing AI2O3 were grown in Teflon jackets (split along top surface) rather than in indium jackets.

Figure 4. Cylindrical test specimens of polycrystalline methane hydrate and mixed sediment aggregates grown by methods described in the text. Left: Sample of pure methane hydrate (CH4»5.89H20), 2.54 cm x 11.5 cm, with uniform grain size of -300 microns and approximately 28% porosity. The encapsulating indium jacket in which the sample was grown has been split and peeled back from the sample. Right: Pure methane hydrate + sediment aggregates fabricated with pre-specified characteristics. Samples are 2.54 cm by 10 - 11.5 cm. Sample at left is methane hydrate + pure quartz sand, varying the proportion of sand in each quarter of the sample. Middle and right samples are mixtures of pure methane hydrate and 1.25 ± 0.25 mm black AI2O3 particulate. The two samples containing AI2O3 were grown in Teflon jackets (split along top surface) rather than in indium jackets.

The resulting material is a cohesive aggregate of uniformly fine, equant, white grains of pure methane hydrate with grain clusters of 250 ± 50 (J.m grain size (Fig. 4). Samples that were initially packed to 40% porosity will contain 29 ± 1% porosity after full reaction due to the volume increase accompanying conversion of the ice to gas hydrate. Purity of the final material and efficiency of the synthesis are highly dependent on adhering to certain key procedures, discussed below. Careful measurements of gas released from fully reacted samples show that the gas hydrate produced at the prescribed test conditions has a highly-reproducible composition of CH4*wH20, where n = 5.89 ± 0.01.

2.2. Methane hydrate + sediment (± seawater) aggregates

Test specimens of methane hydrate + sediment mixtures are easily produced by pre-mixing or layering sediments with granular ice in each reaction vessel prior to admission of CH4 gas. With the static growth method, there is no migration of either H20 or sediment in pre-mixed samples, so samples can be "customized" to any desired mixture or layering sequence. To date, samples have been successfully made with pure methane hydrate inter-layered with discrete horizons of particulate material with grain size as fine as 50 |xm (using SiC) and as coarse as 2 mm (using AI2O3), with no discernable change in the layering during synthesis (Fig. 4). Alternatively, sediments can be homogeneously mixed with seed ice before synthesis to produce a final product that remains fully mixed.

Saturating the gas hydrate or hydrate/sediment samples with seawater (or fresh water) immediately following synthesis is possible but somewhat problematic, in that it can be difficult to assess the extent and distribution of any secondary gas hydrate growth resulting from the added water. A simple method for saturation involves first depressurizing the sample to P-T conditions that are slightly warmer than the ice point but close to equilibrium conditions (for instance, 275 K and 3 MPa Pcm), then admitting highly pressurized seawater through the upper port of the sample such that the pressurized head space over the water pushes the water into the sample and then gravitationally floods the available pore space. Use of the same hydrate-forming gas to pressurize the water prevents possible contamination of the sample due to entrainment or dissolution of the pressurizing gas within the water. Monitoring the weight changes of both the sample and H20 reservoir enables determination of the mass of H20 incorporated into the sample. A more optimal procedure involves a flow-through design to facilitate full saturation of all available pore volume with the added water, easily determined by flowing the water through the base of the sample and observing its emergence from the top end of the sample.

Was this article helpful?

0 0

Post a comment