Document Sample
					Proceedings of the 7th International Conference on Gas Hydrates (ICGH 2011),
Edinburgh, Scotland, United Kingdom, July 17-21, 2011.

                    HYDRATE-BEARING SANDS

                         William Waite*, Peter M. Bratton, David H. Mason
                                      U.S. Geological Survey
                                      384 Woods Hole Road
                                Woods Hole, Massachusetts, 02543

       Naturally occurring hydrate-bearing sands often behave as though methane hydrate is acting as a
       load-bearing member of the sediment. Mimicking this behavior in laboratory samples with
       methane hydrate likely requires forming hydrate from methane dissolved in water. To hasten this
       formation process, we initially form hydrate in a free-gas-limited system, then form additional
       hydrate by circulating methane-supersaturated water through the sample. Though the dissolved-
       phase formation process can theoretically be enhanced by increasing the pore pressure and flow
       rate and lowering the sample temperature, a more fundamental concern is preventing clogs
       resulting from inadvertent methane bubble formation in the circulation lines. Clog prevention
       requires careful temperature control throughout the circulation loop.

        Keywords: gas hydrate, hydrate formation, dissolved phase

NOMENCLATURE                                               phase [7]. This has been accomplished extensively
cc Cubic centimeters                                       in the laboratory using Tetrahydrofuran [8-10].
φ U.S. Sieve size designation (phi)                        Because the solubility of methane in water is low,
                                                           however [11, 12], hydrate formation purely from
INTRODUCTION                                               methane dissolved in water flowing through
Forming gas hydrate, a crystalline solid, in               porous material is a slow process and has not been
sediment alters the host sediment’s mechanical             as frequently attempted [13, 14].
properties. In coarse sediment, where gas hydrate          To accelerate the hydrate-formation process, Priest
tends to form in existing pore space, the impact of        et al. [1] formed methane hydrate from a limited
hydrate formation depends not only on how much             supply of methane bubbles in an otherwise water-
gas hydrate forms, but also on where it forms              saturated sample. This method required ~7 days,
within the pore space [1-3].                               and while this is faster than the several weeks
Hydrate-rich, coarse-grained sediments are                 required for the purely dissolved-phase formation
thought to be of particular value as an energy             of Spangenberg et al. [13], the Priest et al. [1]
resource [4], and these high-hydrate-content               method is limited to pore-space hydrate saturations
sediments are generally thought to behave as if the        below ~40%. Above this level, there is too much
hydrate is a load-bearing member of the sediment           free gas initially to fully transform into hydrate.
[5, 6]. To be most relevant for constraining               We propose to combine the Priest et al. [1] and
mechanical properties in the field, laboratory-            Spangenberg et al. [13] methods. By initiating
formed, hydrate-rich sands should exhibit the same         hydrate growth using the Priest et al. [1] method,
behavior.                                                  we establish an initial hydrate saturation relatively
Load-bearing hydrate systems are thought to occur          quickly. To increase the hydrate saturation, we
when hydrate forms in the absence of a free gas            then adopt the Spangenberg et al. [13] approach

    Corresponding author. Phone: 1-508-457-2346, Fax: 1-508-457-2310, Email:
and circulate methane-rich water through the           respectively). Temperatures are measured via
sample. The flow of water through a hydrate-           thermocouples on the outer surface of the interface
bearing sediment has also been shown by Ebinuma        chamber, in the growth chamber bath, and inside
et al. [15] to migrate grain-cementing hydrate into    the growth chamber near the sample’s top and
load-bearing configurations.                           bottom. Star-oddi DST micro autonomous sensors
While promising in theory, the practical               measure pressure and temperature within the
application of this approach is prone to hydrate       sample itself near the sample’s top and bottom.
clog formation in the inlet and outlet lines, which
                                                       Each sensor is 8.2 mm in diameter, 26 mm long.
limits the capacity for additional hydrate growth in
the sample. Here we examine the formation
process in greater detail, offering suggestions for    PROCEDURE
optimizing hydrate formation rates.                    To begin an experiment, the 5-cm diameter, 240
                                                       cubic centimeter (cc) sample vessel is loaded with
APPARATUS                                              damp F110 quartz sand. To limit migration of sand
We form methane gas hydrate using a flow loop          out of the chamber, the sand is sieved to remove
(Fig. 1). A Polyscience circulator controls the        the < 90 µm grain size fraction (3.5φ sieve), and
growth chamber bath temperature. An Isco pump          61 µm mesh-size screens with 119 µm mesh-size
is used to maintain a baseline pressure in the         screen backings are placed over the inlet and outlet
system. A Quizix pump is used to maintain a            ports. Assuming a target porosity of 36%, the sand
constant water flow rate through the system. We        is pre-moistened to have ~18% pore-space water
use a glass-bead-filled interface chamber to           saturation. The sand is spooned into the sample
saturate the circulating water with methane. With      chamber and tamped to create ~ 1cm-thick layers.
the exception of the Isco and Quizix pumps, the        The Star-oddi sensors are placed at measured
system is housed in a fume hood.                       depths near the bottom and top of the sample. The
Line pressures are measured by both pumps and          thermocouples are driven into the damp-sand
by pressure sensors in the growth chamber’s inlet      sample as the top endcap is seated on the growth
and outlet lines (1.6 and 2.8 mm-inner diameter,       pressure vessel.
                                                       The pressure-temperature path for an experiment
                                                       is shown in Figure 2. The pressure vessel is first
                                                       placed in the temperature-controlled bath at 16°C,
                                                       evacuated, and charged to ~4.2 MPa with methane
                                                       gas. This methane pressure is calculated to provide
                                                       ~20% pore-space hydrate saturation and methane-
                                                       saturate the water that will eventually fill the
                                                       remaining pore space.
                                                       Gas pressure is allowed to stabilize overnight
                                                       before the system is pressurized to 11.7 MPa by
                                                       injecting water with the Isco pump at 0.3 cc/min.
                                                       Approximately a third of the injected water is
                                                       introduced at the base of the sample before
                                                       finishing the pressurization process by injecting
 Figure 1 Flow loop schematic. Water absorbs           from the flow inlet at the top of the sample. This
 methane at room temperature T1 (~23°C) as the         two-sided injection is intended to retain free gas
 water drips past ~250 µm-diameter glass beads         bubbles within the sample’s interior, rather than
 in a methane head space in the gas water              press gas to one end of the sample [E. Rees,
 interface chamber. Methane-rich water then            personal communication], and takes ~ 3.5 hours to
 flows into the growth chamber, where hydrate          complete. Beginning with damp, rather than dry,
 can accumulate in a sand pack held at                 sand yields well-distributed gas bubbles upon
 temperature T2 (9 – 11.5°C) by the growth bath        water injection [18].
 controller. While the pressure control pump           To form hydrate, the sample is cooled to ~4.2°C
 holds the sample outlet pressure at 11.7 MPa,         and held at 11.7 MPa by adding water as needed
 the flow control pump cycles water through the        by the Isco pump. The rate of water consumption
 system at a constant rate between 0.1 and 1
                                                       dwindles to < 0.01 cc/day within ~120 hours.
                                                       After 120 hours, a heater tape attached to the
                                                         Water exits the bottom of the interface chamber
                                                         and passes to the top of the hydrate growth
                                                         chamber, which is nearly submerged in the
                                                         temperature-controlled bath. Circulating water
                                                         cools as it passes through the sample chamber,
                                                         then exits through the heater-tape-warmed drain
                                                         line and completing the circuit to the pumps. By
                                                         leaving the inlet line and sample chamber top
                                                         exposed and out of the hydrate-stability field, we
                                                         seek to reduce hydrate formation in the inlet line
                                                         and top endcap of the growth chamber.
                                                         To conclude an experiment, flow loop valves are
                                                         shut so the sample is in communication only with
                                                         the Isco pump. The sample bath temperature is
                                                         raised to 16°C, 1.5°C above the hydrate stability
                                                         temperature at 11.7 MPa. As hydrate dissociates,
                                                         the Isco pump records the pore water volume
                                                         change at constant pressure.

Figure 2 Pressure-Temperature conditions during          RESULTS
hydrate formation. A: Initial gas injection, ~4.2        In each test, the Isco pump records the volume of
MPa, 16°C. B: Pressurization with injected water,        water injected during the initial hydrate formation,
11.7 MPa, 16°C. C: Initial hydrate formation from        prior to initiating circulatory flow. This represents
gas bubbles in growth chamber, 11.7 MPa, 4.2°C.
                                                         the volume change associated with methane gas in
D: Initial circulation of water to verify clog-free
                                                         the sample converting to methane hydrate. The
flow, 11.7 MPa, 12.5°C. E: Growth chamber
conditions for circulating methane-rich water            volume of water injected is equal, within our
through the hydrate-bearing sand pack, 11.7 MPa,         uncertainties, to the volume of gas and water taken
9-11.5°C (blue circle).      Gas/Water Interface         on by the Isco during hydrate dissociation. This
chamber held at 11.7 MPa, 23°C (orange circle).          balance indicates the circulating methane-rich
Hydrate phase boundary data from [16, 17].               water does not appear to generate a measurable
                                                         volume of new hydrate. Instead, the circulating
growth chamber flow outlet line is turned on, and        water frequently forms hydrate clogs in either the
the sample is warmed to ~12.5°C before water             inlet or outlet lines around the growth chamber.
flow through the sample is initiated. At this            Figure 3 illustrates how small the supersaturation
temperature, circulating water tends to dissolve         ratios (the methane saturation at the interface
small blockages that formed in the inlet and outlet      divided by the methane saturation at the top of the
lines during the initial hydrate synthesis. Once         sample) can be in our system while still resulting
flow is established, the sample is cooled to the         in flow-line clogs.
target temperature of 9 to 11.5°C.
Water circulates through the system at a constant        DISCUSSION
rate of 0.05 to 1.0 cc/min (Fig. 1). Prior to entering   Eliminating flow-line clogging due to hydrate
the growth chamber, water passes over ~250µm-            formation from methane bubbles is a priority.
diameter glass beads in a 100 cc methane-charged         Even in the absence of bubbles, however, choosing
headspace in the interface chamber. The interface        a sample temperature and flow rate requires
chamber is nominally held at 11.7 MPa at a room          balancing competing factors. Here we discuss
temperature of 23°C.                                     bubble mitigation strategies, then present
When circulating water passes through the                considerations for choosing flow rate and sample
interface chamber, the process of dripping through       temperature conditions.
the glass beads exposes a relatively large surface
area of water to the pressurized methane in the          Bubble Mitigation
interface headspace, facilitating uptake of methane      Spangenberg et al. [13] use a relatively high
by the water [M. Rydzy, personal communication].         supersaturation ratio (Fig. 3, green square), and
                                                         though they experienced a clog initially, the clog
                                                        Upon leaving the interface chamber, however, the
                                                        water sits in a precarious equilibrium. If the
                                                        circulating water cools below 13.5°C prior to
                                                        entering the sample, methane can come out of
                                                        solution to form hydrate and clog the line (Fig. 4).
                                                        If instead the circulating water warms at all
                                                        relative to the interface chamber temperature of
                                                        23°C (T1 in Fig. 4), methane can come out of
                                                        solution as a free gas. Bubbles formed in this way
                                                        can be swept to the sample inlet, potentially
                                                        becoming trapped in the 61 or 119 µm mesh-size
                                                        screens. These screens are likely sites for rapid
                                                        hydrate formation from trapped methane bubbles.
                                                        In our system, bubble formation is likely triggered
                                                        in a 50-cm length of metal pipe connecting the
                                                        interface chamber outlet to the growth chamber
                                                        inlet (see Fig. 1). This pipe crosses the heat-
 Figure 3         Relation between methane              venting region of the bath temperature controller
 supersaturation, flow rate, and clog formation.        (lower right in Fig. 1). The exhaust from the bath
 Clogging occurs for nearly all but the lowest          cooler elevates the temperature along the tubing
 supersaturation ratios, which are given by the         length by 3°C relative to the temperature at the
 methane solubility in the interface chamber            interface. To minimize the time water spends in
 divided by that in the growth chamber (see Fig.        this length of pipe, we use a pipe with only a 1.6
 3). Spangenberg et al. [13] (green square)             mm internal diameter. Even so, water still spends
 initially experienced a clog, which did not            one minute or more in the metal pipe depending on
 return     once     removed.     Their    high         flow rate. The trade off for using a small internal-
 supersaturation ratio indicates clog mitigation        diameter pipe is that this provides an efficient
 can be handled without limiting the system to a        geometry for transferring heat from the warm
 low supersaturation ratio.                             exhaust air to the circulating water. Heat shielding
was removed by heating and did not reappear once        and additional insulation will be required to avoid
the sample had again cooled to the target
temperature. The saturation ratio itself is therefore
not likely to be the variable controlling the onset
of clogging in our system. A more reasonable
explanation is bubble formation, with subsequent
rapid hydrate formation.
Figure 4 illustrates the path circulating water takes
in terms of methane solubility. The peaked curved
(blue) denotes the equilibrium number of moles of
methane per kilogram of solution that water can
hold in the dissolved phase (data from Duan et al.
[19]). The solubility peak occurs at 14.5°C, the
phase boundary for methane hydrate at 11.7 MPa.
At temperatures cooler that 14.5°C, hydrate is
stable and excess methane in the circulating water
can come out of solution to form additional
hydrate. For temperatures warmer than 14.5°C,
                                                         Figure 4 Flow loop conditions (black line)
excess methane can come out of solution to form
                                                         relative to the methane solubility in water at
free gas bubbles.                                        11.7 MPa (blue curve). The solubility
The closer the interface chamber is in temperature       difference between the interface (red circle)
(T1 in Fig. 4) to the hydrate stability temperature,     and growth chamber (green circle) is the
the more methane can be absorbed by the water.           methane available for hydrate formation.
bubble formation in this line.

Flow Rate and Sample Temperature
Flow rate and sample temperature control how
much methane is available for hydrate formation,
and how much time the circulating water is in
contact with the sample and can accomplish the
transfer of methane from the dissolved phase to
Figure 5 displays a family of curves showing how
many cubic cm of hydrate could form per day in a
system held at 11.7 MPa, with an interface
chamber temperature of 23°C. The curves assume
1% efficiency in converting dissolved-phase
methane to hydrate. The formation rates observed
by Spangenberg et al. [13] suggest efficiencies
may be 1-10%.                                           Figure 5 Expected hydrate formation volumes
Given the solubility curve shape (Fig. 4), the          per day as a function of sample temperature
methane content in the circulating water can be         and flow rate. The supersaturation ratio ranges
increased by cooling the interface to nearly the        from 1.02 at 12°C to 1.48 at 6°C, but hydrate
hydrate stability temperature. Correspondingly,         formation is assumed to utilize only 1% of the
lowering the sample temperature reduces the             available methane circulating through the
methane concentration that water can hold in the        sample. This efficiency estimate is crudely
presence      of     hydrate.    Increasing     this    based on early results from Spangenberg et al.
                                                        [13]. The interface chamber is assumed to be at
supersaturation difference between the interface
                                                        23°C, 11.7 MPa.
and growth chambers (Fig. 4) is what increases the
amount of methane available per cc of circulating      of hydrate-saturated marine sands. Because of the
water.                                                 low solubility of methane in water, dissolved-
Supersaturation can also be enhanced by elevating      phase formation techniques are likely to be slow.
the pore pressure. Solubility is only weakly           Initiating hydrate growth using less time-
dependent on pressure in the presence of hydrate,      consuming formation techniques, such as the gas-
but the solubility curve for temperatures above the    limited technique of Priest et al. [1] can be an
hydrate equilibrium temperature shifts upward          attractive approach for reducing the volume of
above the hydrate stability temperature as pressure    hydrate formation that must be accomplished from
increases [20], increasing the equilibrium methane     dissolved phase methane.
content in water for a given interface chamber         To enhance the productivity of dissolved-phase
temperature.                                           hydrate formation, methane should be combined
As Figure 5 indicates, elevating the flow rate         with water at the highest attainable pressure, near
elevates the hydrate production by providing           the hydrate stability temperature. Hydrate growth
                                                       should then be accomplished close to 0°C to
additional water to the sample. It is not clear,
                                                       maximize the methane available for hydrate
however, at which flow rate this added
                                                       formation. Flow rates of ~1 cc/min have been
productivity is offset by the diminished time the      successful, but lower rates may be necessary at
circulating water spends in contact with hydrate in    higher hydrate saturations [Spangenberg, personal
the growth chamber. Elevated flow rates may also       communication, 2010].       For a given system
promote      channeled     flow,    which    would     geometry, less efficient pressure, temperature and
significantly limit the spatial and temporal           flow rate parameters me be required in order to
exposure of the circulating water to hydrate in the    avoid bubble formation and subsequent clogging.
growth chamber.                                        The flow line leading from the point of highest
                                                       methane saturation (the interface chamber in this
CONCLUSION                                             work) into the growth chamber is particularly
Hydrate formation from methane dissolved in            vulnerable to bubble formation and must not be
water is thought to result in the load-bearing pore-   allowed to warm above the temperature at which
space hydrate distribution most relevant to studies    methane was saturated into the water. Bubbles
formed in this line can collect at the growth               porous media. Marine Geology, 2000. 164: p.
chamber entrance and rapidly form the small                 69-77.
amount of hydrate required to block the flow line.    8.    Lee, J.Y., T.S. Yun, J.C. Santamarina, and C.
                                                            Ruppel,       Observations      related     to
ACKNOWLEDGMENTS                                             tetrahydrofuran and methane hydrates for
U.S. Geological Survey contributions were                   laboratory studies of hydrate-bearing
supported in part by the Gas Hydrate Project of the         sediments.      Geochemistry       Geophysics
U.S. Geological Survey's Coastal and Marine                 Geosystems,       2007.     8(6):     Q06003,
Geology Program, as well as through the                     doi:10.1029/2006GC001531, 10 p.
Interagency Agreement DE-FE0002911 between
                                                      9.    Pearson, C., J. Murphy, and H. R., Acoustic
the USGS Gas Hydrates Project and the U.S.
                                                            and resistivity measurements on rock samples
Department of Energy's Methane Hydrates R&D
Program. Any use of trade names is for descriptive          containing      Tetrahydrofuran      hydrates:
purposes only and does not imply endorsement by             laborotory analogues to natural gas hydrate
the U.S. Government.                                        deposits. Journal of Geophysical Research,
                                                            1986. 91(B14): p. 14132-14138.
REFERENCES                                            10.   Yun, T.S., J.C. Santamarina, and C. Ruppel,
                                                            Mechanical properties of sand, silt, and clay
1.   Priest, J.A., E.V.L. Rees, and C.R.I. Clayton,         containing tetrahydrofuran hydrate. Journal
     Influence of gas hydrate morphology on the             of Geophysical Research, 2007. 112: B04106,
     seismic velocities of sands. Journal of                doi:10.1029/2006JB004484, 17 p.
     Geophysical Research, 2009. 114(B11):            11.   Lu, W., I.M. Chou, and R.C. Burruss,
     B11205, doi: 10.1029/2009jb006284, 13 p.               Determination of methane concentrations in
2.   Dai, J., F. Snyder, D. Gillespie, A.                   water in equilibrium with sI methane hydrate
     Koesoemadinata, and N. Dutta, Exploration              in the absence of a vapor phase by in situ
     for gas hydrates in the deepwater, northern            Raman      spectroscopy.     Geochimica     et
     Gulf of Mexico: Part I. A seismic approach             Cosmochimica Acta, 2008. 72: p. 412-422.
     based on geologic model, inversion, and rock     12.   Duan, Z.H. and S.D. Mao, A thermodynamic
     physics principles. Marine and Petroleum               model for calculating methane solubility,
     Geology, 2008. 25: p. 830-844.                         density and gas phase composition of
3.   Lee, J.Y., F.M. Francisca, J.C. Santamarina,           methane-bearing aqueous fluids from 273 to
     and C. Ruppel, Parametric study of the                 523 K and from 1 to 2000 bar. Geochimica et
     physical properties of hydrate-bearing sand,           Cosmochimica Acta, 2006. 70(13): p. 3369-
     silt, and clay sediments: 2. Small-strain              3386.
     mechanical       properties.     Journal    of   13.   Spangenberg, E., J. Kulenkampff, R.
     Geophysical Research, 2010. 115(B11):                  Naumann, and J. Erzinger, Pore space
     B11105, doi: 10.1029/2009jb006670, 11p.                hydrate formation in a glass bead sample
4.   Boswell, R. and T.S. Collett, Current                  from     methane     dissolved    in    water.
     perspectives on gas hydrate resources.                 Geophysical Research Letters, 2005. 32:
     Energy and Environmental Science, 2011: p.             L24301, doi:10.1029/2005GL024107, 4 p.
     doi: 10.1039/C0EE00203H, 10p.                    14.   Spangenberg, E., B. Beeskow-Strauch, M.
5.   Kleinberg, R.L. and J. Dai, Estimation of the          Luzi, R. Naumann, and J.M. Schicks. The
     mechanical properties of natural gas hydrate           process of hydrate formation in clastic
     deposits from petrophysical measurements,              sediments and its impact on their physical
     Paper OTC 17205. Offshore Technology                   properties. in 6th International Conference
     Conference, 2005: p. 9 p.                              on Gas Hydrates. 2008. Vancouver, Canada,
6.   Lee, M.W. and W.F. Waite, Estimating pore-             Paper 5648, 8 p.
     space gas hydrate saturations from well-log      15.   Ebinuma, T., K. Suzuki, J. Nagao, H. Oyama,
     acoustic data. Geochemistry Geophysics                 and H. Narita. Ultrasonic Wave Velocities
     Geosystems,       2008.      9(7):    Q07008,          Associated with Formation and Dissociation
     doi:10.1029/2008GC002081, 8p.                          of Methane Hydrate in Artificial Sandy
7.   Buffett, B. and O. Zatsepina, Formation of             Sediment (OTC 19260). in Offshore
     gas hydrate from dissolved gas in natural              Technology Conference. 2008. Houston,
                                                            Texas, U.S.A.
16. Thakore, J.L. and G.D. Holder, Solid-vapor
    azeotropes in hydrate-forming systems.
    Industrial   &     Engineering     Chemistry
    Research, 1987. 26: p. 462 - 469.
17. De Roo, J.L., C.J. Peters, R.N. Lichtenthaler,
    and G.A.M. Diepen, Occurrence of methane
    hydrate in saturated and unsaturated
    solutions of sodium chloride and water In
    dependence of temperature and pressure.
    AIChE Journal, 1983. 29(4): p. 651-657.
18. Domenico, S.N., Effect of Brine-Gas Mixture
    on Velocity in an Unconsolidated Sand
    Reservoir. Geophysics, 1976. 41(5): p. 882-
19. Duan, Z. Methane clathrate hydrate - Liquid
    equilibrium calculations. 2006; Available
    from:                   http://www.geochem-
20. Servio, P. and P. Englezos, Measurement of
    dissolved methane in water in equilibrium
    with its hydrate. Journal of Chemical and
    Engineering Data, 2002. 47(1): p. 87-90.

Shared By: