|IODP Proceedings Volume contents Search|
|Expedition reports Research results Supplementary material Drilling maps Expedition bibliography|
Expedition 311 had the most ambitious pressure coring and onboard pressure core analysis program ever attempted in the history of ocean drilling. Pressure cores retrieved at in situ pressures were used to determine gas hydrate quantity, using degassing techniques and mass balance calculations, and gas hydrate distribution, using nondestructive measurement of the physical properties of the cores at in situ pressures. Large improvements in temperature control over previous expeditions (e.g., Leg 204; Tréhu, Bohrmann, Rack, Torres, et al., 2003) made the recovery and analysis of pressure cores more practical.
Pressure cores were collected using the IODP PCS, the HYACINTH Fugro Pressure Corer (FPC), and the HYACE Rotary Corer (HRC) (see "Description and operation of pressure coring systems"). After a pressure core was retrieved, initial nondestructive measurements were made to characterize the core, determine the core length, and identify massive gas hydrate (see "Nondestructive measurements on pressure cores"). Following X-ray imaging, PCS cores were degassed on board (see "Degassing experiments") to determine total gas composition and quantity in sediments (see "Estimating the abundance of gas hydrate and free gas"). The FPC cores were subsampled and degassed on shore at the Pacific Geoscience Centre of the Geological Survey of Canada immediately following Expedition 311. All cores had gamma ray density measurements made on them while undergoing degassing to document gas evolution, gas hydrate dissociation, or other changes in the core (see "Measurements on pressure coring system cores"). Following degassing, all pressure cores were X-rayed a final time and the released gas volume, X-ray images, and density scans guided subsampling for IW, physical properties, and other related analyses.
Pressure coring is crucial for understanding the concentrations of gas hydrate and free methane gas in marine sediments, their nature and distribution, and their effect on the intrinsic properties of the sediment. Methane and other components of natural gas in deep sediment may be present in three phases:
However, reliable data on methane concentrations are impossible to obtain from conventional coring techniques because conventional cores recovered from ocean depth often release large volumes of gas during recovery (Wallace et al., 2000; Paull and Ussler, 2000). Natural gas solubility decreases significantly as pressure decreases during the recovery of cores to the surface, and any gas volume measurements made on conventional cores will lead to gross underestimates of the in situ natural gas concentrations.
The only way to directly determine the in situ concentrations of natural gas in the subseafloor is to retrieve cores that are sealed immediately after the coring process and recovered to the surface without any loss of the constituents. To achieve this objective, the core must be sealed in an autoclave that is able to maintain the hydrostatic pressure at the coring depth when brought to the surface. This was the concept behind the original PCS, and it has proven to be an essential tool for estimating in situ gas concentrations (Dickens et al., 1997, 2000a, 2000b; Milkov et al., 2004).
Although the PCS is very effective at obtaining samples that are suitable for overall gas concentration analysis, it was not designed to be used for other types of analyses that might reveal the physical structure of gas or gas hydrate in the core. It is also not possible to transfer or sample the PCS core without releasing the pressure. To enable a more comprehensive investigation of gas hydrate–bearing sediments, a more recent program, HYACINTH, has developed not only the next generation of pressure coring tools but has initiated the development of techniques to nondestructively analyze the cores and to take subsamples for microbiological, chemical, and physical property analyses at in situ pressures.
The PCS is a downhole tool designed to recover a 1 m long sediment core with a diameter of 4.32 cm at in situ pressure up to a maximum of 69 MPa (Pettigrew, 1992; Graber et al., 2002). The pressure autoclave consists of an inner core barrel, which ideally collects a 1465 cm3 sediment core, and an outer chamber, which holds 2964 cm3 of seawater/drilling fluids (Fig. F17). Prior to Expedition 311, the PCS was successfully used to study in situ gases in gas-rich and gas hydrate–bearing sediments during Legs 164 on the Blake Ridge (Paull, Matsumoto, Wallace, et al., 1996; Dickens et al., 1997), 201 on the Peru margin (Dickens et al., 2003), and 204 on Hydrate Ridge (Tréhu, Bohrmann, Rack, Torres, et al., 2003; Milkov et al., 2004). In the course of these ODP legs, degassing technology has improved continuously and further modifications were made for Expedition 311 to optimize the control and monitoring of the PCS degassing experiments. Most importantly, we replaced the steel outer and inner barrels of the PCS with aluminum with a maximum working pressure of 250 bar so that recovered sediment could be investigated by the GeoTek pressure multisensor core logger (MSCL-P) X-ray system before and after depressurization (see "Measurements on pressure coring system cores"). We used a one-dimensional vertical gamma ray density scanner to evaluate the distribution of sediment, gas hydrate, and gas voids in the recovered PCS core and to monitor changes in the course of the degassing experiment (see "Nondestructive measurements on pressure cores"). In addition, the manifold was equipped with an additional valve for the collection and analysis of fluids, and the volume of fluid expelled from the PCS was monitored during degassing.
PCS operations and core flow included the following steps. The PCS tool was assembled in and on top of the core technician shop and deployed as during Leg 204 (Tréhu, Bohrmann, Rack, Torres, et al., 2003). When a core was retrieved during Expedition 311, it was immediately inserted into an ice shuck in the moonpool for 20–30 min (see "Improvement of temperature control") to counteract any warming during the wireline trip. The cooled PCS autoclave was separated from the rest of the tool on the rig floor and delivered to a refrigerated van dedicated to HYACINTH logging located on top of the core technician shop. After X-ray imaging, the PCS autoclave was moved to another refrigerated van situated on top of the lab stack for PCS degassing experiments, either using a winch rigged on the porch outside the downhole tools laboratory or simply hand-carried up the stairs.
When degassing experiments were completed, the PCS autoclave was moved back to the core technician shop for core removal. Most of the PCS autoclaves were X-ray imaged again in the HYACINTH van before the inner core barrel was removed from the autoclave. Before removing the core, the water in the autoclave was carefully collected by opening the top valves all the way and the ball valve very slightly to allow water in the inner and outer barrel to flow out of the ball valve. This volume of water was measured for consideration in mass balance calculations from the depressurization experiments. The inner barrel was removed from the rest of the PCS autoclave and taken back to the HYACINTH van for X-ray analysis. This second X-ray analysis was necessary to obtain X-ray images of the entire core because steel components inside the outer core barrel obstruct X-ray imaging of the PCS autoclave below a core depth of 51 cm (Fig. F17). Final extrusion of the core into a half-liner took place in the core tech shop using a metal plug and broom handle or hydraulic pump (as dictated by the sediment stiffness). The core was given to the IODP curator and, with the aid of the X-ray images and gamma ray density profiles (see "Measurements on pressure coring system cores"), samples were taken for analysis of IW, physical properties, and dissolved gases.
Two types of wireline pressure coring tools were developed in the HYACE/HYACINTH programs: a percussion corer and a rotary corer, which were designed to cut and recover core in a wide range of lithologies where gas hydrate–bearing formations might exist (Schultheiss et al., 2005). Both tools have been designed for use with the same IODP bottom-hole assembly (BHA) as the PCS (i.e., the APC/XCB BHA). The HYACINTH pressure coring system was used successfully during Leg 204 to recover gas hydrate and surrounding sediments (Tréhu, Bohrmann, Rack, Torres, et al., 2003).
The design and operation of the HYACINTH tools differs in four significant respects from that of the PCS:
The HYACINTH percussion corer was developed by Fugro Engineers BV and is known as the Fugro Pressure Corer (Fig. F18). The FPC uses a water hammer, driven by the circulating fluid pumped down the drill pipe, to drive the core barrel into the sediment up to 1 m ahead of the drill bit. The core diameter is 57 mm (liner outer diameter = 63 mm). On completion of coring, the drill string is lifted to extract the core barrel from the sediment. Once the core barrel is free from the sediment, the wireline pulls the core barrel liner containing the core into the autoclave. A specially designed flapper valve seals the bottom end of the autoclave after the core has been retrieved. The FPC is designed to retain a pressure of up to 25 MPa. It is suitable for use with unlithified sediments ranging from soft through stiff clays to sandy or gravelly material. In soft sediments it acts like a push corer prior to the hammer mechanism becoming active. It has operated effectively in sediments with shear strengths exceeding 500 kPa.
The HYACINTH rotary corer was developed by the Technical University of Berlin and the Technical University of Clausthal and is known as the HYACE Rotary Corer (Fig. F18). HYACE was the name of the original development program. The HRC uses an inverse Moineau motor driven by the circulating fluid pumped down the drill pipe to rotate the cutting shoe up to 1 m ahead of the roller cone bit. A narrow kerf, dry auger design cutting shoe with polycrystalline diamond cutting elements is used on the HRC. This design allows the core to enter into the inner barrel before any flushing fluid can contaminate the material being cored. The core diameter is 51 mm (liner outer diameter is 56 mm). On completion of coring, the tool is lifted off the bottom with the drill string and then the core is retracted into the autoclave by pulling in on the wireline in a similar manner to the FPC. The pressure is sealed by a specially designed flapper valve. The HRC is designed to retain a pressure of up to 25 MPa and was primarily designed for use in sampling lithified sediment or rock. However, in practice we have found that the HRC can also sample much softer formations very effectively, presumably acting as a push corer with minimal rotation.
As during Leg 204, the HRC and the FPC were prepared and assembled on tool trestles located on the port side of the piperacker. The normal tool assembly area above the core technician shop was used for PCS tool assembly and was impacted by the presence of the 20 ft HYACINTH van. Stands of drill pipe normally used from the port side were moved to the starboard side to reduce disruption to the tool preparations.
Both tools followed similar operational procedures on the rig floor. They were initially transferred from the piperacker working area into the vertical position. To achieve this, a tugger line from the derrick was attached to the upper end of the tool while the base of the tool was lowered onto the piperacker skate using the port side racker crane. The tool was then hauled into a vertical position using the tugger line and lowered into the rig floor shuck as the strongbacks were removed by hand. Finally, the tool was deployed in the open drill string that was then closed, and the tools were lowered on the wireline while pumping and rotating.
When the tools were recovered to the rig floor, they were placed into the ice water–filled shuck in the moonpool for 30 min, similar to the recovery of the PCS. Once removed from the ice shuck, both the FPC and the HRC followed a reverse procedure back to the trestles on the piperacker, including replacing the strongbacks. Autoclaves were removed from the tools in a timely manner (<15 min) and placed in the HYACINTH cold van. It was once thought that additional ice baths might be necessary to rechill the autoclaves at this point; however, temperature data from the autoclave data loggers proved this to be unnecessary.
Between Leg 204 and Expedition 311, the HYACINTH transfer and analysis systems were redesigned and integrated to fit inside a 20 ft refrigerated van, and procedures differ significantly from those described in Tréhu, Bohrmann, Rack, Torres, et al. (2003) and Schultheiss et al. (2005). To remove the core from the pressure corer autoclave, the autoclave was connected to the manipulator/shear transfer chamber with quick-clamps (Fig. F19B) and then pressure balanced with the autoclave before opening the ball valves. The "technical" end of the pressure core containing the piston and other components was captured by a catcher on the end of the manipulator, and the full core was withdrawn from the autoclave into the shear transfer chamber, the ball valves closed, and the autoclave removed from the system.
The manipulator/shear transfer chamber (STC), now containing the core at full in situ pressure, was attached to the GeoTek MSCL-P (see "Pressure multisensor core logger measurements on HYACINTH cores"), pressures were balanced, and ball valves were opened (Fig. F19D). The core was pushed and pulled through the sensors using the manipulator under computer control. Once the analyses were completed, the core was withdrawn to the cutting position (Fig. F19E) and the core was cut free from the manipulator portion of the core system with the shear blades. A storage chamber was then attached to the manipulator/STC and pressures balanced. The core was pushed into this storage chamber (Fig. F19F) for storage at in situ pressure and temperature conditions (5°–7°C) for shore-based analyses.
At the first two sites (U1328 and U1329) cored we used seawater as the pressurizing medium, as had been done on previous expeditions, but at the remaining three sites we used freshwater, which is much less corrosive for long-term storage. The freshwater pressurizing fluid was spiked with fluorescein (1–10 mg/L), and samples of pressurizing fluid were taken when each pressure core was stored so that shore-based investigators might monitor any infiltration of the pressurizing fluid into the core.
To study the properties of gas hydrate–bearing sediments from sediment cores, an ideal core would retain the in situ effective pressures, the hydrostatic pressure, and the temperature. It is currently only practical to retain the hydrostatic pressure in a coring tool. During Expedition 311, we also made an effort to improve temperature control during the recovery, handling, and analysis of pressure cores.
Pressure cores rarely (if ever) arrive in the laboratory at in situ pressure (Dickens et al, 2003; Tréhu, Bohrmann, Rack, Torres, et al., 2003). The recovery pressure is generally below the in situ pressure, but occasionally the recovery pressure has been higher when there is gas hydrate present. The main causes of pressure loss are
Volume changes from differential pressure occur mainly from the initial compression of compliant components (O-rings, etc.) as the tool seals, though a small component is caused by the volume expansion of the tool itself as the pressure on the outside falls with respect to the inside pressure.
Even cores recovered at substantially below in situ pressure and those that have been outside the GHSZ for a significant period of time are still of value if the corer sealed soon after coring, capturing all enclosed methane and other core constituents. If pressure losses can be attributed mainly to tool volume changes due to pressure and temperature, although the methane may have changed phase during the retrieval process, the total quantity of methane will remain unchanged and the in situ phases can be calculated. However, cores that have been outside gas hydrate stability cannot be used to investigate the nature and distribution of gas hydrate in the core with confidence using nondestructive techniques (P-wave velocity, gamma ray density, and X-ray imaging). The difficulty arises in recognizing the difference between pressure losses due to late sealing of a pressure coring tool versus losses caused by pressure and temperature changes. Thus, the operation of the tools has become an area of ongoing investigation.
To minimize the reduction in pressure caused by differential expansion from temperature and pressure effects, the HRC and FPC systems contain a gas accumulator that is normally set at 80%–90% of the anticipated in situ pressure. This allows the tool to expand slightly without any significant change in pressure, keeping the pressure high and minimizing the chances of the core moving out of the GHSZ. PCS deployments during Expedition 311 generally recovered cores with 50%–60% of the in situ pressure. The PCS could benefit from a gas-filled pressure accumulator connected to the autoclave.
None of the pressure coring tools have any active temperature control and, hence, the best that can be achieved is to minimize or reverse any adverse rises in temperature during the complete coring and handling processes. To achieve this, we recovered the core to the rig floor on the wireline "as fast as practically possible;" normally at a speed of 100 m/min but up to 250 m/min. After breaking the corer out of the pipe, it was quickly inserted through a rathole in the rig floor into a newly designed, vertical, 2.5 m deep, insulated ice water–filled shuck suspended in the moonpool (Fig. F20). This ice shuck is deep enough to quickly cool the autoclave containing the core, which is at the bottom of the tools, and was filled by a chute from an ice machine located below the drill floor in the subsea shack. After examining the temperature records for the first few pressure corer deployments, a 30 min ice soak before tool removal and breakdown was deemed optimal. During this chilling period, the next rig floor operation/tool deployment was performed and, hence, the cooling time had little or no impact on drilling activities. The autoclave temperature was 0°–3°C after chilling, and the autoclave slowly warmed during autoclave removal at ambient temperature before it was moved into the refrigerated van. The thermal mass of the core, surrounding fluid, and autoclave was sufficient to keep the core from warming more than a few degrees in 10–20 min. Thereafter, further analyses on the autoclave were conducted in dedicated cold vans.
The inclusion of dedicated cold vans for analysis of pressurized cores was a major change from previous expeditions. During Leg 204, degassing of the PCS cores took place in iced cylinders in the hard rock laboratory, next to the thin section laboratory, while the transfer of HYACINTH cores took place on the walkway above the catwalk with ice bags being used (unsatisfactorily) for cooling. Degassing and logging of the HYACINTH cores took place in very warm conditions in the lab stack hold, with frequent interruptions, despite insulating foam around the pressure chamber, to take the core back into the adjacent cold reefers in an effort to prevent rapid dissociation of the gas hydrate. During Expedition 311, all of these operations took place in two temperature-controlled (5°–7°C) 20 ft containers, one dedicated to PCS degassing experiments and one used for transferring and logging the HYACINTH cores in the MSCL-P, as well as for X-raying the PCS cores before and after degassing. With this procedure, we maximized the chances of keeping the core within the GHSZ during its journey from the seafloor to the laboratory.
Although pressure cores are particularly valuable for providing accurate methane volumes for gas hydrate concentration calculations, nondestructive measurements made before or during the depressurization process can provide additional information on the nature and distribution of gas hydrate within the sediment and rare data on near in situ physical properties of gas hydrate–bearing sediments. X-ray images of the pressure cores show the overall structure of the core and gas hydrate within them (as well as contributing to the core length estimate), GRA provides accurate densities of sediment/gas hydrate structures, and measurement of P-wave velocity on undisturbed gas hydrate–bearing sediments at in situ pressure provides acoustic parameters valuable for analysis of seismic data. Two measurement systems were used during Expedition 311 to collect data on pressure cores: the GeoTek MSCL-P, used on cores under full recovered pressure; and the GeoTek vertical multisensor core logger (MSCL-V), used during degassing experiments.
Efforts during Expedition 311 were the first attempt at taking nondestructive measurements on PCS cores while still in the autoclave under pressure. The aluminum core barrels (specially fabricated for this expedition) allowed X-ray analysis of the PCS cores. Being able to "see" the core prior to degassing enabled the original length (and hence volume) of the core to be measured, which is critical in the calculation of gas hydrate content. The density measurements taken during degassing using the MSCL-V provided information similar to that collected on HYACINTH cores during Leg 204, where low-density, potentially gas hydrate–bearing layers could be monitored as the core was depressurized to observe gas hydrate dissociation and gas evolution.
The PCS autoclave was brought to the HYACINTH logging van for X-ray scanning after it was removed from the tool body. The top of the autoclave was mated to the end of an unpressurized HYACINTH manipulator, allowing the MSCL-P software and manipulator to push the PCS autoclave through the X-ray imaging system. Unfortunately, the metal body of the PCS caused distortion of the X-ray images (Fig. F21). As the PCS was moved past the image intensifier, the character of this classic "S"-distortion changed (Fig. F21), probably as a result of the moving steel interacting with the magnetic fields in the image intensifier. An X-ray montage from 0 to 51 cm relative to the top of the PCS inner barrel was created (normally in 0.5 cm increments) for each PCS core. The lower half of the barrel was completely obscured by a steel sleeve (Fig. F17B).
Once the PCS autoclaves were moved to the PCS degassing van atop the lab stack, they were placed in the MSCL-V (Fig. F22) (Tréhu, Bohrmann, Rack, Torres, et al., 2003). The MSCL-V accommodates cores vertically, and the sensor cluster moves up and down along the stationary core. The gamma ray source and detectors are the same as those used on the MSCL-P and IODP MST (see "Multisensor track"), and calibration was performed in a similar fashion (see "Pressure multisensor core logger measurements on HYACINTH cores"). The GRA of an aluminum calibration sample of varying, known thickness was measured within a water-filled PCS autoclave to provide the density calibration. The PCS autoclaves were always oriented the same way in the MSCL-V, with the transducer port facing forward, and gamma ray attenuation for the PCS autoclaves, filled with water, was measured so that the data could be corrected as a function of vertical position (Fig. F17).
The core densities measured in the upper half of the PCS autoclave had an estimated error of ±0.05 g/cm3, but because the lower half of the autoclave outer barrel contains a spring and other steel sleeves, the density of the core in the lower half of the PCS autoclave could not be determined as accurately. There was also an unexplained interaction between the PCS and the gamma ray attenuation sensor at 30–55 cm core depth, causing a lowering of the count rate that varied over time in this area (Fig. F17). However, the primary use of the tool was to look at density differences that occurred during degassing, and, hence, most of the data were simply plotted as differential density (the initial density profile subtracted from each of the subsequent profiles) to observe the evolution and migration of gas within the core barrel during depressurization.
The MSCL-P (Fig. F23) is an automated measurement system for the collection of acoustic P-wave velocity, GRA, and X-ray image data on HYACINTH pressure cores under pressures up to 25 MPa. The MSCL-P pressure chamber is constructed of aluminum and contains an internal set of ultrasonic transducers. X-ray and gamma ray sources and detectors are situated outside of the pressure chamber. The system moves pressurized HYACINTH cores incrementally past these sensors, under computer control, with a positional precision of better than 1 mm, allowing detailed gamma ray density and acoustic velocity profiles to be obtained rapidly and automatically along the core as well as creating automated full-core X-ray montages. The manipulator mechanism ensures that the core does not rotate during the linear translation.
Core logging under pressure using the MSCL-P is, in principle, very similar to core logging with the IODP MST or a standard GeoTek MSCL. One exception is the increased distance and varied material between the sensors and the core. Sensors are separated from the core by the plastic liner, the pressurizing fluid (seawater), and, in the case of the gamma and X-ray sensors, the aluminum pressure chamber. To calibrate for measurements of acoustic velocity and gamma ray density, similar techniques are used to those developed for the MST and MSCL (see "Multisensor track"), which use distilled water and aluminum as standards. During logging of pressure cores, the inner liner is assumed to have a constant diameter because it cannot be directly measured under pressure.
Gamma ray density was measured using a 137Cs source and NaI detector very similar to those used on the MST (see "Multisensor track"). Errors are proportional to the square root of the total counts (generally ~5000 cps), giving a density precision of 2%. Calibration of the gamma ray density measurements was performed by measuring the intensity of the gamma ray beam through a stepped aluminum bar of varying thickness sitting centrally in a core liner filled with and surrounded by saltwater of known salinity. This calibration procedure, using aluminum and water, provides a good approximation for a water-saturated sediment (minerals and water) and has proven to be an excellent calibration protocol for determining density from GRA. Separate calibrations were performed for FPC and HRC liners, and no effect was seen with increasing pressure.
Ultrasonic P-wave velocity (VP) was measured using two 500 kHz acoustic transducers mounted inside the pressure chamber, perpendicular to the core and the gamma ray beam. Traveltimes were measured with a precision of 50 ns, and the error associated with the velocity was ±3 m/s, assuming a core thickness of ~6 cm. To calibrate VP , the total P-wave traveltime was measured when both the core liner and the pressure chamber were filled with water of known velocity (from temperature, pressure, and salinity). Changes in traveltime as a function of pressure were also measured (up to 25 MPa). The measured variation in VP with pressure was close to the theoretical variation for water, and therefore the traveltimes in the liner material were essentially constant with changing pressure (as was found during Leg 204; Tréhu, Bohrmann, Rack, Torres, et al., 2003).
X-ray images were obtained using a linear X-ray device consisting of a lead-shielded microfocal X-ray source, phosphor image intensifier, and digital camera. An aluminum compensator minimizes the intensity variations that are caused when illuminating round objects. With the geometrical arrangement and typical X-ray spot size of ~8–12 µm used, the intrinsic spatial resolution of the images is ~150 µm. All final core images were obtained by creating montages from a series of area images taken along the core. The normal spatial interval used for the final image was 0.5 cm, which creates a relatively flat image along the core without any apparent significant spherical distortion. However, an unexpected electromagnetic distortion in the image intensifier, similar to but less intense than that observed with the PCS (Fig. F21), limited our ability to create perfectly smoothed montages. The X-ray images were not density calibrated because the GRA measurements give higher accuracy than the polychromatic X-rays. Instead, we varied the X-ray energy and power levels to maximize the qualitative resolution of the image in an effort to examine subtle structures within the core. X-ray energies as much as 110 kV were used depending on the density of the cores being measured.
The HYACINTH pressure cores were placed in storage chambers for immediate postcruise analysis and sampling at the Pacific Geoscience Centre of the Geological Survey of Canada, Sidney, British Columbia. The cores underwent one of three different subsampling and analysis protocols. Cores to be degassed were degassed at 4°C, in a similar fashion to the PCS, using the same degassing manifold, MicroGC, and the MSCL-V. Cores to be subsampled and placed into Parr vessels were rapidly depressurized by opening a side valve on the storage chamber, and the core was cut using a hacksaw into 20 cm subsections that would fit inside the Parr vessels. The subsections were placed inside the Parr vessels and repressurized with methane. Other pressure cores were held for specialized subsampling under pressure.
During Expedition 311, the PCS retrieved pressurized sediments for onboard degassing experiments. Controlled release of pressure from the PCS through a manifold permits
Degassing experiments were carried out in the PCS van (Fig. F24) after some time (>1 h) had been allowed for the PCS to equilibrate to ambient van temperature (7°C) and a vertical gamma ray scan had been run to determine the initial density distribution within the PCS core (see "Measurements on pressure coring system cores"). The PCS was connected to a pressure transducer, a helium-flushed degassing manifold for controlled release of pressure, and through the manifold to a bubbling chamber that allows collection of released gas (Fig. F25). We used a liquid leak detector to check the connection between the PCS and the manifold for leakages. Prior to each degassing experiment, we monitored the air quality in the temperature-controlled van and analyzed a blank (i.e., a gas-phase sample taken from the helium-flushed bubbling chamber). If the blank contained traces of O2 and N2, we continued flushing of the degassing system with helium and repeated the blank analysis until all air was removed. For the gas analysis, we used the same Agilent 3000A MicroGC and method as for samples (see below). Thereafter, we carefully released and collected a small volume of gas that was usually extruded together with some water from the outer core barrel (~500 mL) to flush the lines connecting the PCS to the manifold. All water that escaped from the PCS was collected by the sampling port on the manifold, quantified, and subsampled for geochemical analysis. We observed a strong decrease in pressure when water was expelled from the PCS during the initial phase of the degassing experiment. In the following steps, gas was released, collected, and subsampled from the bubbling chamber for quantitative and qualitative analysis as outlined in the degassing protocol below. During the degassing procedure, we ran additional gamma ray scans of the PCS to monitor the evolution of gas voids and pathways. Degassing experiments were terminated when the pressure within the PCS had equilibrated to atmospheric pressure, and <5 mL of gas had exsolved from the core within 3 h. At the end of each experiment, we collected a final gas sample and ran a final gamma ray scan. The degassed core was removed from the outer core barrel, X-rayed, extruded, curated, and sampled for IW chemistry, dissolved gases, and physical properties, including parameters critical for calculating methane concentration (see "Estimating the abundance of gas hydrate and free gas").
Degassing was carried out in incremental steps using the following protocol. First, the manifold was closed with respect to the bubbling chamber and then opened with respect to the PCS. In this manner, a constant volume of gas was allowed to move into the manifold. The pressure inside the manifold was not constant throughout the experiment but equilibrated with the residual pressure of the PCS core. Next, the valve between the manifold and the PCS was closed and the gas inside the manifold was released into the bubbling chamber where its volume expanded as a result of the pressure release.
PCS cores that had strong indications for the presence of gas hydrates were degassed in the following way. During an initial phase, gas and any water that was expelled from the outer core barrel were released in small incremental steps. At this stage degassing caused immediate pressure drops inside the PCS. Once pressure had reached equilibrium conditions for gas hydrate stability, further release of gas caused only relatively small decreases in pressure. The reason for this behavior is twofold:
During the gas hydrate dissociation phase, depressurization was repeatedly stopped to monitor the pressure response and to carry out gamma ray density scans. Once pressure had dropped below equilibrium conditions for gas hydrate stability, we monitored carefully whether further incremental release of gas was followed by pressure increases from gas hydrate dissociation. When pressure decreases indicated the absence of gas hydrate, depressurization was carried on continuously.
During each degassing experiment, the internal PCS pressure was monitored by analog pressure gauges and recorded by digital pressured transducers connected to a computer. However, because of technical problems, digital pressure records were not available for all degassing experiments conducted during Expedition 311. The released gas was collected in a 1 L bubbling chamber consisting of an inverted graduated cylinder and a plexiglass tube filled with a saturated NaCl solution (Fig. F25). The released gas volume was recorded as a function of PCS opening number and pressure. After measuring the volume of collected gas, gas aliquots were sampled from a valve at the top of the inverted cylinder using a syringe. The syringe was thoroughly flushed with gas from the bubbling chamber to minimize dilution of the sample by air that was present in the syringe and valve prior to sampling. After flushing the syringe, one aliquot (6 mL) was taken for immediate analysis of the gas composition (C1, C2, CO2, N2, and O2) using an Agilent 3000A MicroGC equipped with Plot U and molecular sieve columns and a TCD that was located in the PCS van. Another aliquot (10 mL) was taken for further shipboard analysis (e.g., low concentrations of C2 and higher hydrocarbon gases) and for shore-based isotopic analyses. These samples were stored in headspace vials filled with saturated NaCl solution and tightly sealed with untreated blue butyl rubber stoppers and crimp caps. During this expedition, any air measured in gas samples from the PCS was assumed to have been laboratory contamination, and the samples were corrected for air dilution.
Aside from the released gas volume and pressure response inside the PCS, further parameters were documented that are crucial for accurate mass balance calculations and interpretation of the degassing experiments. First, to calculate the quantity of released methane by the ideal gas law, ambient air pressure and air temperature in the PCS van were continuously recorded throughout the experiment using a digital temperature logger inside the PCS van and the barometer on the bridge deck. Second, when pressure is released from the PCS, free gas can evolve in such a way that it forces water out of the core. The total volume of expelled water corresponds to the volume of gas left inside the PCS at the end of depressurization. We accounted for the gas remaining in the PCS by recording the volume of expelled water and assigning the composition of the final gas sample. Third, after degassing was finished, we collected the water remaining in the outer core barrel to account for any additional headspace that might have been present in the PCS. However, we did not include these data in the mass balance calculations presented in this report, because any incomplete recovery of water from the outer core barrel would lead to an overestimation of in situ gas hydrate concentrations. Finally, to account for the total pore water volume, the accurate length of the recovered sediment interval was obtained by an initial gamma ray density scan and the porosity of the sediment was analyzed from solid phase samples taken at the end of the experiment. For mass balance calculations, we used the length obtained prior to degassing (reported as recovered length in each site chapter) rather than the length obtained after degassing from the extruded core (reported as curated length in each site chapter), because any loss of sediment from the inner core barrel during degassing and core extrusion would lead to an overestimation of in situ gas hydrate concentrations. In addition, we took samples for HS analysis from the extruded sediment to account for dissolved methane in each PCS core (see "Organic geochemistry").
Pressure core degassing experiments allow characterization of the total amount of methane that is present in the PCS as free and dissolved gas phases at shipboard pressure and temperature conditions using the equation
ntot CH4 = nfree CH4 + ndiss CH4, (5)
We calculated the quantity of free methane from the measured total volume of released gas, the molar fraction of methane in the gas phase known from continuous GC analysis during the degassing experiment, ambient air temperature, and pressure using the ideal gas law
nfree CH4 = M x patm x V/(R x T), (6)
The total volume of released gas resulted from the gas volume released into the bubbling chamber, the volume of expelled water, which was assumed to leave a corresponding gas void in the PCS, and from the dead volume of the degassing system (261 mL), which was filled with gas released from the PCS at the end of the degassing experiment.
The amount of dissolved methane was analyzed from sediment samples taken at the end of the experiment using the headspace gas analysis protocol (see "Organic geochemistry") and calculated by
ndiss CH4 = [M x patm x VH/(R x T x x VS)] x Vpw , (7)
Pore volume (Vpw) is given by
Vpw = x rPCS2 x lPCS x , (8)
The porosity used in mass balance calculations was the mean of samples taken from the PCS as well as from APC and XCB cores taken at similar depths from the same site.
The total methane concentration (cCH4) in the pore space of the PCS core is given by the total amount of methane and the pore water volume inside the PCS core:
cCH4 = ntotCH4/Vpw . (9)
To account for the amount of methane that was enclosed in the PCS as free gas or gas hydrate at in situ conditions, we considered the variation of methane solubility with temperature, pressure, and salinity. We calculated methane solubility according to Xu (2002, 2004). Based on the in situ methane saturation, we determined the concentration of methane that could have been dissolved in the pore water volume within the PCS (Vpw). If the pore water was oversaturated with methane at in situ conditions, the concentration of nondissolved methane was obtained by subtracting the concentration of dissolved methane cdissCH4, in situ (i.e., methane saturation) from the total methane concentration in the PCS (cCH4). The amount of nondissolved methane in the PCS was calculated from its concentration and the pore water volume inside the PCS:
nfreeCH4, in situ = (cCH4 – cdissCH4, in situ) x Vpw . (10)
If in situ pressure and temperature were outside the gas hydrate stability field, nondissolved methane was considered to be free gas and the volume of the free gas phase was calculated based on the ideal gas law. If the in situ pressure and temperature indicated gas hydrate stability, nondissolved methane was considered to be gas hydrate and its volume (VGH) was calculated based on and the molecular weight (mGH) and density (GH) of gas hydrate using
VGH = nfree CH4, in situ x mGH/GH, (11)
Finally, we calculated how much water would have been released and freshened the pore water inside the PCS if the predicted amount of gas hydrate had decomposed during the degassing experiment, assuming that 108 g of water is released per mole of decomposed gas hydrate.
Degassing experiments at Site U1329 were the first ones performed during Expedition 311. They revealed that an initial gas volume of 500 mL needs to be released before all helium is removed from the hoses that connect the PCS port to the bubbling chamber. The hoses and manifold hold a dead volume of 261 mL. Methane that was released within the initial gas volume was included when calculating the total amount of methane released from the core. However, the methane concentration of the helium-diluted initial gas volume was excluded when the average composition of the released gas phase was determined.