Hydrogen sulfide–rich hydrates and saline fluids in the continental
margin of South Australia
P.K. Swart
Marine Geology and Geophysics, Rosenstiel School of Marine and Atmospheric Sciences, 4600 Rickenbacker Causeway,
Miami, Florida 33149, USA
U.G. Wortmann Geomar Research Center for Marine Geosciences, Wischofstrasse 1-3, D-24148 Kiel, Germany
R.M. Mitterer Department of Geoscience, University of Texas at Dallas, P.O. Box 830688, Richardson, Texas 75083-0688, USA
M.J. Malone Ocean Drilling Program, Texas A&M University, 1000 Discovery Drive, College Station, Texas 77845, USA
P.L. Smart Department of Geography, University of Bristol, University Road, Bristol BS8 1SS, UK
D.A. Feary* Australian Geological Survey Organisation, GPO Box 378, Canberra, ACT 2601, Australia
A.C. Hine Department of Marine Science, University of South Florida, 1407th Avenue South, St. Petersburg, Florida 33701, USA
ABSTRACT
During the drilling of the southern Australian continental margin (Leg 182 of the Ocean
Drilling Program), fluids with unusually high salinities (to 106‰) were encountered in Miocene
to Pleistocene sediments. At three sites (1127, 1129, and 1131), high contents of H2S (to 15%),
CH4 (50%), and CO2 (70%) were also encountered. These levels of H2S are the highest yet reported during the history of either the Deep Sea Drilling Project or the Ocean Drilling Program.
The high concentrations of H2S and CH4 are associated with anomalous Na+/Cl– ratios in the
pore waters. Although hydrates were not recovered, and despite the shallow water depth of these
sites (200–400 m) and relative warm bottom water temperatures (11–14 °C), we believe that
these sites possess disseminated H2S-dominated hydrates. This contention is supported by calculations using the measured gas concentrations and temperatures of the cores, and depths of
recovery. High concentrations of H2S necessary for the formation of hydrates under these conditions were provided by the abundant SO42– caused by the high salinities of the pore fluids, and
the high concentrations of organic material. One hypothesis for the origin of these fluids is that
they were formed on the adjacent continental shelf during previous lowstands of sea level and
were forced into the sediments under the influence of hydrostatic head.
Keywords: hydrates, methane, hydrogen sulfide, chloride, pore water.
INTRODUCTION
The southern margin of the Australian continent is a divergent, passive continental margin
that formed during the extension and rifting that
led to the separation of Australia and Antarctica
in the Cretaceous (Willcox et al., 1988; Davies
et al., 1989). The modern continental shelf is sediment starved because, although there are high
rates of sedimentation on the adjacent margin
(Feary et al., 2000), there is little accumulation on
the shelf and late Neogene rocks are exposed at
the surface (Feary and James, 1995). The shallow
depth of the shelf means that it was subaerially
exposed numerous times during the Pleistocene.
Nine sites were drilled during Ocean Drilling
Program (ODP) Leg 182, in water depths from
200 to 3900 m (Fig. 1). The water depths, depths
of penetration, and other relevant geochemical
data are shown in Table 1. At all sites whole-round
samples, varying between 5 and 10 cm in length,
were taken at intervals of ~9.5 m. These samples
were squeezed to obtain interstitial pore fluids and
analyzed for salinity, Cl–, Na+, Mg2+, Ca2+, Sr2+,
alkalinity, SO 42–, and NH+4 using standard ODP
procedures (Gieskes et al., 1991). In addition, gas
pockets and headspace gases were analyzed to de-
termine concentrations of H2S, CH4, C2H6, C3H8,
and CO2 (Kvenvolden and McDonald, 1986). The
geochemical results together with other information on the sedimentology, stratigraphy, and
operations are included in Feary et al. (2000).
Although some of the deeper water sites contained noncarbonate minerals, the three sites that
are the focus of this paper consisted mainly of
various forms of calcium carbonate with minor
concentrations of dolomite (Feary et al., 2000).
RESULTS
A significant discovery made during the drilling is the presence of a brine that reached a maximum salinity of 106‰, and is present in and
underlying seven sites (the exceptions being Site
1128, drilled in a water depth of 3884 m, and Site
1133). The brine is present at relatively shallow
subsurface depths at the deeper sites, 1130, 1134
and 1126, whereas at the shallower sites (1127,
1129, 1131, and 1132) the maximum salinities
are not encountered until deeper than 400 m below
seafloor (mbsf) (Fig. 2A). At three sites (1129,
1131, and 1127; the eastern transect) the Na+/Cl–
ratio of the pore fluids significantly exceeds the
same ratio in seawater (0.81) (Fig. 2B). The
elevated Na+/Cl– ratios do not increase with depth,
but reach a maximum approximately coincident
with the maximum concentrations of CH4 and
Figure 1. Location map
showing position of sites
drilled during Ocean
Drilling Program Leg 182
on Eucla margin. Lower
panel shows hypothetical
cross section through
Sites 1127, 1131, and
1129 (Feary and James,
1995).
*Present Address: National Research Council,
2101 Constitution Avenue NW, Washington, D.C.
20418, USA.
Geology; November 2000; v. 28; no. 11; p. 1039–1042; 6 figures; 2 tables.
1039
H2S. The gas concentrations reach values as high
as 16% (H2S) and 50% (CH4) at Site 1131
(Table 2). It is well established that gas concentrations obtained in this manner are minimums
(Paull et al., 1996). At Sites 1129, 1131, and 1127
(Fig. 1), salinity increased with depth, reaching a
maximum value of ~100‰ at Site 1127, farthest
away from the margin, and 90‰ at Site 1129,
closest to the shelf. A similar pattern in salinity
was observed at Sites 1130 and 1132 (western
transect), although the concentrations of CH4 and
H2S remained low and there were only small
increases in the Na+/Cl– ratio.
All three sites along the eastern transect are
characterized by high alkalinity values and low
SO 42–/Cl– ratios. The only site at which SO 42– is
Na+/ Cl –
Cl – (mM)
500
1000
1500
2000
0.80
0.90
1.00
1.10
1.20
0
100
Depth (mbsf)
200
300
400
500
1127
1131
1129
A
B
600
Figure 2. A: Concentrations of Cl– with respect to depth from eastern transect sites (mbsf,
meters below seafloor). B: Na+/Cl– ratios from eastern transect sites. Vertical shaded bar
shows seawater ratio.
1040
completely removed is Site 1127 between 87
and 177 mbsf. At Site 1131 and 1129, minima in
the SO 42–/Cl– ratio are reached between 240 and
300 mbsf (Fig. 3). Below these depths the
SO 42–/Cl– ratio increases as a result of the presence of the high-salinity brine (Fig. 3). The western transect did not show significant decreases in
the SO 42–/Cl– ratio.
ORIGIN OF THE
HIGH-SALINITY FLUIDS
The maximum salinity is reached at approximately the same depth below the sea surface at all
sites. This is evident in a contour plot of the
concentration of Cl– at Sites 1127, 1131, and 1129
(Fig. 4). At the deeper sites the maximum salinities are observed at very shallow subbottom
depths, the maximum salinity occurring at a depth
of only 41 mbsf at Site 1126 and 43 mbsf at Site
1134. This common depth of the brine below the
sea surface suggests that the fluids were probably
emplaced in sediments of differing ages under the
influence of a hydrostatic head. Although the
origin of the fluids has not yet been established, a
probable explanation is that the fluids formed
during the Pleistocene when the Eucla shelf was
subaerially exposed numerous times. We suggest
that large hypersaline lagoons developed on the
shelf in the arid climate. By virtue of the hydrostatic head, fluids were forced into underlying
strata and out onto the adjacent continental slope.
During the following sea-level rise the shelf
flooded and sedimentation started anew on the
slope. Ions then diffused upward as sediment was
deposited, the profiles of individual species being
modified by diagenetic reactions. This model is
supported by the Cl– distribution (Fig. 4), which
suggests that the top of the brine has a common
depth below sea level and therefore crosscuts
sequence boundaries. The penetration of fluids
derived from aquifers within the continental shelf
for considerable distances has been documented
at numerous sites drilled in the Deep Sea Drilling
Project (DSDP) and ODP. Typically these fluids
can be identified as a result of decreases in salinity
such as that documented at Site 241 off the east
African coast (Gieskes, 1974). At other sites, such
as those drilled during Leg 40 in the Angola basin
(Sotelo and Gieskes, 1978) and Leg 112 off Peru
(Kastner et al., 1990), the presence of highly
saline fluids derived from the penetration of fluids
GEOLOGY, November 2000
The most common guest gas in hydrates is
methane (CH4). Hydrates dominated by CH4
form typically in waters colder than 10 °C and at
depths >800 m. In contrast, hydrates in which
H2S is dominant can form as shallow as 500 m at
the same temperature (Fig. 5). Although hydrates
have been drilled and recovered from many locations during coring by the DSDP and the ODP
(Shipley and Didyk, 1982; Harrison and Curiale,
1982; Kvenvolden and MacDonald, 1985;
Kvenvolden et al., 1993; Westbrook et al., 1994;
Paull et al., 1996; Kastner et al., 1998), hydrates
frequently decompose during the retrieval of the
core, leaving little evidence that they were once
present other than high concentrations of gases
and the persistence of geochemical anomalies in
the pore waters.
through aquifers containing evaporitic strata have
been identified.
Temperature °C
5
0
10
15
20
25
30
70:30
2
90:10
Pressure (Mpa)
ORIGIN OF HIGH GAS
CONCENTRATIONS AND
GEOCHEMICAL ANOMALIES
We suggest that the geochemical anomalies in
the pore waters and the high concentrations of
CH4 and H2S arise from the presence of disseminated gas hydrates in the sediments. Gas
hydrates are crystalline substances composed
principally of three-dimensional cages of solid
H2O in which various guest gases can enter and
stabilize the structure (Sloan, 1990). Conditions
necessary for the formation of hydrates include
high concentrations of guest gases in addition to
conditions of low temperature and high pressure.
Pore fluids in the Pleistocene portion of the
sediments from the eastern transect have a
Na+/Cl– ratio in excess of that of seawater
(Fig. 2B). Along the eastern transect, the Na+/Cl–
ratio is in excess of unity, while at Sites 1130 and
1132 (the western transect) the Na+/Cl– ratios are
only slightly elevated. The presence of fluids
with Na+/Cl– close to unity would appear to
suggest that some of the brines were involved in
0
4
6
100
Depth (mbsf)
8
100 %
Methane
Figure 3. Excess SO42– in pore
waters from eastern transect
sites. Only site from which all
SO42– was removed was Site
1127.This site also contained
highest concentrations of CH4
(mbsf—meters below sea
floor).
200
300
10
Figure 5. Influence of addition of H2S
on stability of hydrates as calculated
using Colorado School of Mines hydrate (CSMHYD) program (Sloan,
1990). Lines represent various mixtures between pure methane hydrates and hydrates with composition of 70% CH4 and 30% H2S.
Rectangular box represents approximate range of temperatures and
pressures at eastern transect sites.
400
1127
500
–30 –20 –10
1131
1129
0 –30 –20 –10
0 –30 –20 –10
Excess SO42– mM
0
Temperature
Site 1127
Site 1131
Site 1129
0.0
1
2
0
0.4
900
0.6
5
10
15
20
C
25
30
3 km
water bottom
0.2
Two-way traveltim e
NORTH
0
700
1100
2
Pressure (mPa)
SO UTH
o
4
1129
6
1131
8
0.8
900
1100
1.0
1300
1400
Figure 4. Contour map showing Cl– concentrations from eastern transect sites overlying interpretation of seismic line through three sites. Note that Cl– concentrations crosscut seismic
sequences and appear to originate from brine with common depth below sea level. Cl– concentration has been contoured using data from three sites and Kriging method.
GEOLOGY, November 2000
1127
10
Figure 6. Hydrate stability plot for maximum
H2S concentration encountered at Ocean
Drilling Program Site 1131. Dashed line
shows stability of mixture of 80% CH4 and
20% H2S. Dotted lines represent approximate geothermal gradients observed at
eastern transect sites. Solid lines superimposed on dotted lines represent zones of
maximum gas concentration at each site.
1041
the dissolution of salt deposits, perhaps lower in
the section. However, that the Na+/Cl– ratio does
not increase with depth and is therefore not related to the increase in salinity (Fig. 2) suggests
that the origin of the increase is not related to the
origin of the brine. While Na+/Cl– ratios in excess
of seawater values have been measured at other
ODP sites, the association of the maximum
anomaly with the zone of inferred hydrate stability
suggests an association between hydrate formation and the Na+/Cl– ratio. Other hydrate locations where data are available on the Na+ and Cl–
concentrations, such as the Cascadia subduction
zone, also exhibit elevated Na+/Cl– ratios (Westbrook et al., 1994). In addition, a calculation of
the Na+ concentration using the data presented by
Paull et al. (1996) for sediments containing
hydrates from the Blake Ridge also indicates
elevated Na+/Cl– ratios. One possible mechanism
of formation for these elevated Na+/Cl– ratios is
that as water is used in the formation of hydrates,
salts are left behind, perhaps reaching the point at
which NaCl crystallizes. Once the salt is removed
from solution it becomes decoupled from the
fluid, and subsequent melting of the hydrates can
dissolve the halite and produce fluids with
Na+/Cl– ratios close to unity. This process would
necessitate that the salinity locally would be ~350
and may therefore be unrealistic. An alternative
explanation is that Cl– is preferentially incorporated in the hydrate structure relative to Na+.
Support for this latter hypothesis is provided by
the fact that the Na+/Cl– ratios actually exceed
unity at Site 1131, reaching values as high as 1.1.
The presence of gas hydrates is supported by
modeling results. Using the Colorado School of
Mines hydrate (CSMHYD) model of gas hydrate formation (Sloan, 1990), the temperatures
measured during the cruise, the measured salinities, and the maximum concentrations of gases
measured at each of the three eastern sites (Feary
et al., 2000) (Table 2), it is possible to calculate the
depth at which hydrates are stable. These calculations, shown in Figure 6, indicate that hydrates are
stable at Site 1127 over the depth range where the
maximum H2S and CH4 were measured. At
Site 1131, where the highest concentrations of
CH4 and H2S were measured, the temperature of
the interval from which the gases were collected
is at present too warm for hydrates to be stable.
However, a comparison of the concentrations of
gases in the samples collected from gas pockets,
which were collected only over a small range of
depths, with head space samples indicates that
similar high concentrations of H2S might be
present in gas pockets at 100 mbsf, within the hydrate stability zone. At Site 1129, the concentrations of gas are too low to support the presence of
hydrates. It is perhaps significant that the Na+/Cl–
anomaly at Site 1127 corresponds with the zone
where thermodynamic calculations support present
hydrate formation, whereas at the other two sites
(1129 and 1131) the Na+/Cl– anomaly is signifi1042
cantly shallower than the zones of maximum gas
composition. This may indicate that there may
have been enhanced hydrate formation previously
and that as the sediments became buried the
hydrates became unstable and dissociated. It is
significant that, while the measured gas concentrations indicate that gas hydrate formation is possible in these sediments, it is universally recognized that gas concentrations measured using the
techniques employed by the ODP are a minimum
estimate of the actual values in the core. It has
been suggested that the concentrations may be incorrect by at least an order of magnitude (Paull
et al., 1996; Dickens et al., 1997). Hence it is possible that hydrates are present at all of the sites in
the eastern transect.
A further question is why no bottom simulating
reflector (BSR) was detected at these sites. Although the occurrence of hydrates in the absence
of a BSR is not unprecedented (Paull et al., 1996),
the absence of a BSR at the eastern sites is probably a result of the varying amounts of H2S at each
site as well as high pore-water salinities, which are
known to inhibit the formation of hydrates (Sloan,
1990), and the disseminated nature of the hydrate.
The only other documented finding of H2Srich hydrates was from Site 892 off the coast of
Oregon (Kastner et al., 1998). These hydrates
were associated with a maximum of 10% H2S,
compared to 16% in this study. The explanation
for the higher levels of H2S in this study arises
from the association with the high-salinity fluids,
which provide a greater supply of SO 42– for the
oxidation of organic material. We suggest that
H2S-dominated hydrates may be more common
than previously realized, particularly in continental margin settings where there is an abundant
source of organic material and the possibility of
saline-rich fluids. This has significant implications because relatively small changes in sea level
can result in the release of large quantities of CH4
and H2S into the atmosphere.
ACKNOWLEDGMENTS
We thank the crew and technicians of the JOIDES
Resolution for their invaluable help in obtaining the
samples for this research. We received support from: a
U.S. Science Advisory Committee grant (Swart), the
Natural Environment Research Council (Swart), and
the German Science Foundation (Wortmann). This
paper benefited from reviews by J. Gieskes and A. Mills.
REFERENCES CITED
Davies, H., Clarke, J.A., Stagg, H.M.J., Shafik, S.,
McGowan, B., and Willcox, J.B., 1989, Maastrichtian and younger sediments from the Great
Australian Bight: Australian Bureau of Mineral
Resources Report 288, 40 p.
Dickens, G.R., Paull, C.K., Wallace, P., and the ODP
Leg 164 Scientific Party, 1997, Direct measurement of in situ methane quantities in a large gashydrate reservoir: Nature, v. 385, p. 426–428.
Feary, D.A., and James, N.P., 1995, Cenozoic biogenic
mounds and buried Miocene (?) barrier reef on a
predominantly cool-water carbonate continental
margin, Eucla Basin, western Great Australian
Bight: Geology, v. 23, p. 427–430.
Printed in U.S.A.
Feary, D.A., Hine, A.C., Malone, M.J., et al., 2000, Proceedings of the Ocean Drilling Program, Initial
reports, Volume 182: College Station, Texas,
Ocean Drilling Program, CD-ROM.
Gieskes, J.M., 1974, Interstitial water studies, Leg 25,
in Schlich, R., Initial reports of the Deep Sea
Drilling Project, Volume 25: Washington, D.C.,
U.S. Government Printing Office, p. 361–394.
Gieskes, J.M., Gamo, T., and Brumsack, H., 1991,
Chemical methods for interstitial water analysis
aboard JOIDES Resolution: Ocean Drilling Program Technical Note 15: College Station, Texas,
Ocean Drilling Program, 60 p.
Harrison, W.E., and Curiale, J.A., 1982, Gas hydrates in
sediments of Holes 497 and 498A, Deep Sea
Drilling Project Leg 67, in Auboin, J., von
Huene, R., et al., Initial reports of the Deep Sea
Drilling Project, Volume 67: Washington, D.C.,
U.S. Government Printing Office, p. 591–594.
Kastner, M., Elderfield, H., Martin, J., Suess, E.,
Kvenvolden, K., and Garrison, R., 1990, Diagenesis and interstitial-water chemistry at the
Peruvian continental margin-major constituents
and strontium isotopes, in Suess, E., and von
Huene, R., Proceedings of the Ocean Drilling Program, Scientific results, Volume 112: College Station, Texas, Ocean Drilling Program, p. 413–440.
Kastner, M., Kvenvolden, K., and Lorseson, T.D.,
1998, Chemistry, isotopic composition, and origin of a methane-hydrogen sulfide hydrate at the
Cascadia subduction zone: Earth and Planetary
Science Letters, v. 156, p. 173–183.
Kvenvolden, K.A., and McDonald, T.J., 1986, Gas
hydrates of the middle America trench—Deep
Sea Drilling Project Leg 184, in von Huene, R.,
Auboin, J., et al., Initial reports of the Deep Sea
Drilling Project, Volume 84: Washington, D.C.,
U.S. Government Printing Office, p. 667–682.
Kvenvolden, K.A., and McDonald, T.J., 1986, Organic
geochemistry on the JOIDES Resolution—An
assay: Ocean Drilling Program Technical Note 6:
College Station, Texas, Ocean Drilling Program,
147 p.
Kvenvolden, K.A., Ginsburg, G.D., and Soloviev, V.A.,
1993, Worldwide distribution of subaquatic gas
hydrates: Geo-Marine Letters, v. 13, p. 32–40.
Paull, C.K., Matsumoto, R., Wallace, P.J., et al., 1996,
Proceedings of the Ocean Drilling Program, Initial reports, Volume 164: College Station, Texas,
Ocean Drilling Program, 623 p.
Shipley, T.H., and Didyk, B.M., 1982, Occurrence of
methane hydrate offshore southern Mexico, in
Watkins, J.S., Moore, J.C., et al., Initial reports of
the Deep Sea Drilling Project, Volume 66: Washington, D.C., U.S. Government Printing Office,
p. 547–555.
Sloan, E.D., 1990, Clathrate hydrates of natural gases:
New York, Marcel Dekker, 641 p.
Sotelo, V., and Giekes, J., 1978, Interstitial water studies,
Leg 40: Shipboard results, in Boli, H.M., Ryan
W.B.F., et al., Initial reports of the Deep Sea
Drilling Project, Volume 40: Washington, D.C.,
U.S. Government Printing Office, p. 549–551.
Westbrook, G.K., Carson, B., Musgrave, R.J., et al.,
1994, Proceedings of the Ocean Drilling Program, Initial reports, Volume 146: College Station, Texas, Ocean Drilling Program, 92 p.
Willcox, J., Stagg, H.M.J., Davies, H., et al., 1988, Rig
seismic research cruises 10 & 11: Geology of the
central Great Australian Bight region: Australian
Bureau of Mineral Resources Report 286, 140 p.
Manuscript received March 13, 2000
Revised manuscript received August 3, 2000
Manuscript accepted August 21, 2000
GEOLOGY, November 2000