Absolute values and
downhole trends of CH4 
13C
collected from gas voids in APC and XCB cores have been used by Paull et al. (Chap.
7, this volume) to make inferences on the origin and migration of CH4
on the Blake Ridge. Such interpretations are based on the fundamental assumption
that limited fractionation of carbon isotopes occurs during core recovery
despite an enormous loss of gas (Dickens et al., 1997).
Volumetrically weighted
averages of CH4 13C
for gas samples collected at pressure by the PCS are the same as the CH4
13C
for gas samples collected from gas voids in normal APC and XCB cores. The
relationship holds for PCS cores that contained relatively low and high total
gas volumes. Thus, CH4 13C
data from gas voids appears to be representative of bulk in situ gas
irrespective of the amount of gas lost. Interestingly, our conclusion for
isotope fractionation of CH4 may not extend to the molecular
distribution of hydrocarbon gases. Gas samples collected from the PCS are often
significantly enriched in heavier hydrocarbons (C4 through C7)
when compared to gas samples collected from gas voids in APC and XCB cores (Paull,
Matsumoto, Wallace, et al., 1996).
The isotopic values of the
gas void and PCS gas samples indicate that CH4 in Blake Ridge
sediments is largely microbial in origin (Paull et al., Chap.
7, this volume). Furthermore, the relatively constant 13C
values at depths greater than ~300 mbsf indicate that little or no addition of
thermogenic gas occurs with increasing depth (Fig.
1). The cause of the systematic decreases in methane 13C
values observed at shallower depths is complex and is likely the result of
upward migration of CH4 and CO2 coupled with Rayleigh
fractionation during bacterially mediated CO2 reduction (Paull et
al., Chap. 7, this
volume).
At Sites 994, 995, and
997, the base of gas hydrate stability is present at a depth of ~450 mbsf, and a
bottom simulating reflector (BSR) occurs at approximately this level at Sites
995 and 997. Several lines of geochemical and geophysical evidence indicate that
the BSR in this region of the Blake Ridge is the interface between CH4
hydrate-bearing sediments above and free gas-bearing sediments below (Paull,
Matsumoto, Wallace, et al., 1996; Holbrook et al., 1996; Dickens et al., 1997).
At Sites 995 and 997, no discontinuity is observed in methane 13C
values across the BSR (Fig. 1).
This observation suggests that there is no significant equilibrium carbon
isotopic fractionation in reactions involving CH4 hydrate, CH4-rich
free gas, and dissolved CH4 in pore waters. Such an interpretation is
consistent with studies of both natural and experimentally synthesized CH4
hydrate (Claypool et al., 1985; Pflaum et al., 1986; Sassen and MacDonald,
1997).
The CH4 13C
for individual gas samples collected during degassing of a single PCS core vary
by as much as 9
.
The large range typically reflects an anomalous gas sample that was collected
from a small volume of gas released from the PCS at high pressure. This
observation suggests that kinetic fractionation of carbon isotopes can occur
during rapid degassing of the PCS.
Simple analog degassing
experiments demonstrate that kinetic effects associated with degassing can
result in significant carbon isotope fractionation of CH4. The
magnitude of this fractionation appears to increase with the complexity of the
system that is being degassed. Thus, the experiments involving only pressurized
CH4 show smaller isotope fractionations than do the experiments
involving CH4-saturated water (up to 2).
The carbon isotope fractionations that can occur during PCS degassing may
reflect the complex system inside of the PCS, which can include CH4
gas, CH4 hydrate, CH4-saturated water, and sediment
located in multiple chambers separated by valves.
Dickens et al. (Chap. 11, this volume) have highlighted two additional factors that complicate any interpretation of data from PCS degassing experiments. First, significant quantities of CH4-poor borehole water fill the PCS during deployment and come into contact with the core after it is sealed in the pressurized housing. This leads to dilution of CH4 concentration in interstitial water in the sediment core and, in many cases, decomposition of CH4 hydrate before a degassing experiment begins. Second, degassing experiments were conducted after the PCS had equilibrated in an ice-water bath (0șC). This temperature is significantly lower than in situ values in the formation before core recovery and could lead to additional CH4 hydrate formation before degassing began.
There are two main carbon
isotope fractionation processes during PCS degassing experiments (Fig.
2, Fig. 3, Fig.
4, Fig. 5, Fig.
6, and Fig. 7). The
first and most significant is during the initial degassing of high pressure CH4-poor
air that is trapped inside the PCS during deployment. The second, which is on
the order of 1-2,
occurs during release of larger volumes of CH4. Based on comparison
with the results of the analog degassing experiments, fractionations of this
magnitude can be caused by (1) high-pressure release of headspace gas that is
then bubbled through water, and (2) failure to allow the gas + water ± CH4-hydrate
system inside the PCS to reequilibrate after the large pressure drop associated
with removal of an aliquot of gas. The effects of these processes are probably
amplified when interstitial gas must escape through fine-grained sediment.
Solving the first problem, that of trapped air inside the PCS, would require modifying the deployment procedure and interior design of the PCS so that only water is initially present inside the pressure chamber. However, as demonstrated here, the effects of trapped air are relatively easy to recognize and do not pose a major problem in interpreting PCS gas data. During future experiments with the PCS, the second problem (disequilibrium) can be minimized if gas is released in relatively small aliquots, resulting in small pressure drops, and if the PCS is allowed to reequilibrate at the new pressure before releasing another aliquot of gas. However, during Leg 164, the total time available for conducting a given PCS degassing experiment was limited by the need to prepare the tool for another coring run. A redesigned pressure core sampler that allows samples to be removed from the tool while maintaining pressure would make it possible to degass samples as slowly as is necessary to ensure reequilibration between gas release steps.
