Previous   |    Next

doi:10.2204/iodp.proc.308.215.2012

Expedition 308 drilling, coring, logging, and downhole measurements

During Expedition 308, we drilled to just above the Blue Unit at three locations, Sites U1322–U1324 (Figs. F1, F2). At Site U1323 during the initial LWD phase, overpressured and unconsolidated sediments were encountered, resulting in flow of water and sand up the borehole (shallow-water flow) (see the “Expedition 308 summary” chapter [Expedition 308 Scientists, 2006a]). The well was terminated at ~250 mbsf, and no coring was accomplished at this location. We did core and make in situ pressure measurements at Sites U1322 and U1324.

Stratigraphy

Sawyer et al. (2009) described the stratigraphy of the drilling transect (Fig. F2). The leveed-channel systems above the Blue Unit contain sandstone channels and bounding levees. The Ursa Canyon channel-levee system immediately overlies the Blue Unit, whereas the Southwest Pass Canyon channel-levee system is younger and lies west of Ursa Canyon (Fig. F2). Mudstones above the Southwest Pass Canyon record deposition from leveed channels that lay further to the west, outside of the study area (Fig. F2B). Multiple mass transport deposits (MTDs) are present within the leveed-channel deposits. These MTDs are of the same composition as the mudstone-prone levee deposition.

Site U1324 is composed of lithostratigraphic Units I and II as defined by Expedition 308 (see the “Site U1324” chapter [Expedition 308 Scientists, 2006d]) (Fig. F3A). Unit I extends from 0 to 365 mbsf and is composed of clay and silt; Unit II is composed of interbedded silt, sandstone, and clay and ranges from 365 to 604 mbsf (Fig. F3A). Seismic reflection S40-1324 lies at the boundary between these two lithofacies. Reflections between the seafloor and reflection S40-1324 are generally parallel and continuous and are controlled by minor changes in bulk density and velocity. Below reflection S40-1324, the responses of the seismic, gamma ray, resistivity, velocity, and bulk density logs are more variable. This variability is due in part to the variable lithology (interbedded sandstone, siltstone, and claystone) and in part to some hole washout within the sandstone intervals.

No core was collected at Site U1323; however, LWD logs were collected to 247 mbsf (Fig. F4). Three units were defined (see the “Site U1323” chapter [Expedition 308 Scientists, 2006c]). Logging Unit I extends from the seafloor to 197 mbsf and is interpreted to be dominated by mudstone with several siltstone intervals and two MTDs (Fig. F4). At the top of logging Unit II, gamma ray and resistivity sharply decrease with depth, which is interpreted to record a transition downhole to sandstone. Logging Unit II cannot be correlated to Site U1324 or U1322. In logging Unit III, gamma ray values increase with depth, which suggests a downhole increase in clay content. At the base of logging Unit III (242 mbsf), a high-amplitude reflection correlates with relatively low gamma ray values; we interpret this to indicate sandstone.

Site U1322 is divided into lithostratigraphic Units I (0–125 mbsf) and II (125–238 mbsf) (see the “Site U1322” chapter [Expedition 308 Scientists, 2006b]) (Fig. F5A). These units are dominated by silty claystone, and the majority of the section is composed of MTDs, as illustrated by the white zones in Figure F2 and the purple zones in Figure F5A. MTDs are recorded by increased resistivity, increased bulk density, decreased porosity, and increased shear strength. The base of each MTD is marked by an abrupt decrease in density and increase in porosity with depth (Fig. F5C); this is recorded with a prominent negative (black) seismic reflection (Sawyer et al., 2009).

Eight coeval depositional surfaces were interpreted by Sawyer et al. (2009) (Fig. F2B). The base of the Blue Unit onlaps a regionally extensive condensed section, which contains the extinction events of the planktonic foraminifer Globorotalia flexuosa (70 ka) and the calcareous nannofossil Pontosphaera 1 (~70 ka) (Styzen, 1995; Winker and Booth, 2000). The age of the top of the Blue Unit is unknown. Expedition 308 cored to within ~20 m of the top of the Blue Unit at Sites U1324 and U1322 and recovered sediments younger than 57 ka (see the “Expedition 308 summary” chapter [Expedition 308 Scientists, 2006]). At Site U1324, age markers of 48, 42, 24, 16, and 10 ka were identified (Fig. F2B). At Site U1322, age markers of only 57, 24, 16, and 10 ka were identified. However, the 57 and 24 ka markers are tentative because they were recovered within MTDs. We correlate these horizons across the transect with the exception of the 57, 48, and 42 ka timelines because they were not identified at both Sites U1322 and U1324.

Porosity

The consolidation of sediment with depth at Ursa Basin was studied by Long et al. (2011). At Site U1324, porosity declines from 80% to 54% within the uppermost 50 mbsf (Fig. F3D). Between 50 and 520 mbsf, mudstone porosity gradually declines from 54% to 42%. Finally, between 520 mbsf and the bottom of the hole (612 mbsf), porosity declines from 42% to 37%. In the sandstone- and siltstone-rich zones beneath 300 mbsf, the caliper log has high values, indicating borehole enlargement. High porosities are measured with the LWD logs in these intervals; Sawyer et al. (2009) interpret that the LWD density values are partially recording the open borehole.

At Site U1322, porosity decreases from 80% to 54% within the uppermost 50 mbsf; between 50 and 120 mbsf, porosity gradually declines to 44% (Fig. F5D). Beneath this, porosity is approximately constant. At Site U1322, nearly the entire section is composed of silty claystone (Fig. F5A, F5B).

Mass transport deposits

MTDs are largely transparent in seismic data (Fig. F2). Log and core data demonstrate that MTDs are more dense than over- or underlying undeformed material (Sawyer et al., 2009; Strong, 2009). They are present at all three sites; however, Site U1322 has many more MTDs than Site U1324 (purple bedding in Figs. F3A, F5A). The best example of an MTD is illustrated at Site U1322 between 90 and 125 mbsf (Fig. F5). Sawyer et al. (2009) interpreted that the decreased porosity within MTDs (relative to the bounding sediment) resulted from vertical uniaxial consolidation during burial of a MTD that had been “remolded” during transport as a debris flow. Remolding destroys the soil fabric and deflocculates the material. As a result, the soil has higher compressibility and it compresses to a lower porosity at a given effective stress than it would have if it had not been remolded (Burland, 1990; Chandler, 2000; Mitchell, 1993; Skempton, 1969).

Pressure

A critical goal of Expedition 308 was to measure in situ pore pressure, one of the significant successes of the expedition (Flemings et al., 2008; Long et al., 2008a). We used the Davis-Villinger Temperature-Pressure Probe (DVTPP) (Moore, Taira, Klaus, et al., 2001; Tréhu, Bohrmann, Rack, Torres, et al., 2003) and the temperature/dual pressure (T2P) probe (Flemings et al., 2006) to measure in situ pressure (Fig. F5). Pore pressure penetrometer data require significant interpretation, and different approaches are pursued (Flemings et al., 2008; Long et al., 2007). We describe the results from Sites U1324 and U1322 below. Long et al. (2008a) described the details of each deployment and how the tools were calibrated.

We plot overpressure versus depth beneath seafloor in Figures F3E and F5E. In these plots, pore pressure (u) equals hydrostatic pressure when overpressure (u*) is zero, whereas pore pressure equals overburden stress (σ_v) when overpressure (u*) equals hydrostatic effective stress (σ_vh′). σ_vh′ is commonly referred to as reduced overburden stress (σ_vh′ = σ_v – u_h) and is delineated with a solid black line in each plot (Figs. F3E, F5E). We also use the overpressure ratio [λ* = (u – u_h)/(σ_v – u_h)] to characterize the relationship between pore pressure and overburden stress. When λ* = 0, overpressure equals zero and pore pressure is hydrostatic. When λ* = 1, pore pressure equals overburden stress.

At Site U1324, above 200 mbsf, severe overpressures parallel the reduced overburden stress, and λ* = 0.7 (Fig. F3E). Beneath 200 mbsf, overpressures are low and approximately equal to 1 MPa. Downhole pressures measured during drilling varied between 1.2 and 1.4 MPa beneath 400 mbsf (PWD in Fig. F3E). When sandstones were penetrated, pressures dropped relative to mudstone penetration. We interpret that this pressure drop records flow from the borehole into the permeable formation (circulation loss) and that the downhole pressures are an upper bound to the formation pressure.

The overpressure distribution at Site U1324 is closely tied to lithology. Above 300 mbsf, where overpressure is high, the sediments are composed of monotonous hemipelagic silty claystone (Figs. F2, F3). At 305.7 mbsf, the first sandstone is encountered. This massive fine to medium sandstone is 1.1 m thick and is the coarsest sandstone bed encountered at Site U1324. It underlies the first decline in overpressure with depth, which is recorded at 250 mbsf. This sandstone lies within the eastern levee of Southwest Pass Canyon, which is otherwise dominated by clay and fine silt. It is truncated eastward between Sites U1324 and U1323 by a MTD, extends to the south, and ultimately intersects the channel fill of Southwest Pass Canyon (Sawyer et al., 2009). Its lateral continuity to the north and west is unknown. Beneath this sandstone, where overpressure is constant, silty claystone is interbedded with beds of silt and very fine sandstone (Fig. F3).

At Site U1322, overpressure increases linearly from the seafloor to 200 mbsf (Fig. F5). The sediments at Site U1322 are composed of silty claystone that is similar to the upper 200 m at Site U1324 (Fig. F3). The pore pressure relative to the overburden is slightly less at Site U1322 than at Site U1324 (λ* = 0.6 versus λ* = 0.7). The deepest measurement (marked with a “?” in Fig. F5E) has relatively low pressure (0.25 MPa) and is of high quality. A high-quality measurement occurs when there is significant dissipation over a long time (e.g., 15–30 min) and no tool dislodgement (Long et al., 2008a). The low pressure observed is contrary to the general trend of increasing overpressure with depth present at Site U1322 (Fig. F5E). This low pressure at the base of the hole may be caused by penetration and drainage of fluid from the Blue Unit during industry drilling since the mid-1990s (Flemings et al., 2008). Many of the wells that were drilled in this region penetrated the Blue Unit sandstone, and significant volumes of water and unconsolidated sand were expelled at the seafloor (Alberty, 2000).

Laboratory analysis

An extraordinary amount of whole core was taken during Expedition 308 that allowed for detailed geomechanical analyses (see the “Expedition 308 summary” chapter [Expedition 308 Scientists, 2006]). Grain size analyses were performed (Sawyer et al., 2008) in addition to clay mineralogy analysis (John and Adatte, 2009). Long et al. (2008b) described the compression behavior and the permeability of the Ursa Basin mudstones. Long et al. (2011) showed that the compression behavior observed in the laboratory was similar to the field compression behavior. Schneider (2011) analyzed the Long et al. (2008b) results and quantified an exponential relationship between porosity and permeability. Dugan and Germaine (2009) completed triaxial strength experiments on specimens from Sites U1322 and U1324 and documented internal friction angles from 17° to 29°.

Synthesis through modeling

A range of hydrodynamic modeling has been performed on data from the Ursa system based on the results from Expedition 308 (Binh et al., 2010; Dugan and Stigall, 2010; Flemings et al., 2008; Stigall and Dugan, 2010; Urgeles et al., 2010). See Table T1 for a list of consolidation nomenclature used during Expedition 308. All of these models follow the consolidation equation

where

• u = pressure,
• k = permeability,
• µ = dynamic viscosity, and
• Dσ_v/Dt = the loading rate from sedimentation.

m_v, the uniaxial compressibility, is defined by

where V is the bulk volume. The coefficient of consolidation (c_v), often termed the hydraulic diffusivity, is

Overpressure is generated when the loading rate by sedimentation (Dσ_v/Dt) outpaces the rate at which pressure can dissipate (c_v—²u*). Gibson (1958) presented an analytical solution for Equation 1 and showed that the degree of overpressure is controlled by the dimensionless time factor (T_g):

where m is sedimentation rate and t is total time. T_g > ~1.0 results in significant overpressure (Flemings et al., 2008; Gibson, 1958).

We use both experimental results and in situ observations to describe the porosity (n), permeability (k), compressibility (m_v), and coefficient of consolidation (c_v) for the three characteristic lithofacies encountered: mudstone, MTD, and siltstone (Fig. F6; Table T1).

Permeability as a function of effective stress was explored by Long et al. (2008b) and Schneider (2011). We describe permeability as an exponential function of porosity:

where n is porosity and γ and k_o are empirically derived from experimental work. We performed 29 constant rate of strain (CRS) experiments on Ursa Basin mudstones to determine the permeability of individual core samples at a range of porosities (Long et al., 2008b). From these data, we calculated the harmonic average of the permeabilities at a range of porosities. We used linear regression on these averaged data to determine k_o and γ (Equation 5) for the mudstone lithofacies (Table T1). We assume that the harmonically averaged relationship describes the vertical permeability. We compared mudstones from within MTDs to undeformed samples and could find no distinguishable difference in the permeability-porosity relationship. Thus, we describe the permeability behavior of the MTD lithofacies using the same parameters as those of the mudstone lithofacies. Long et al. (2008b) presented one CRS test on siltstone and we use these measurements to describe siltstone lithofacies permeability (Table T1).

Long et al. (2011) used CRS experiments to determine the compression behavior of mudstones. They described the in situ porosity-effective stress behavior with

where e_o is the void ratio at an effective stress of unity, and σ₀ is the reference stress, which we choose as 1 MPa. Long et al. (2011) find C = –0.0979 and e_o = 0.89 at 1 MPa when units of MPa are used (Table T2). As discussed, at a given effective stress the MTDs are more compressed (lower void ratio) than the undeformed mudstone. We assume for the MTD that the exponent C is the same as that for the mudstone. However, we assume a lower value for e_o such that on a log(e + 1) – log(σ_v′) plot the compression line for the MTD lithofacies is moved to the left half a log cycle (10^0.5) from the compression line of the mudstone lithofacies (Table T2).

The compression behavior of the siltstone lithofacies (Table T2) is based on a single CRS test (Long et al., 2008b). This compression behavior is best described using a conventional soil model where

where e_o is the void ratio at an effective stress of unity and σ₀ is the reference stress, which we choose as 1 MPa. We discuss the behavior of porosity (n), permeability (k), compressibility (m_v), and coefficient of consolidation (c_v) over a stress range from 0.01 to 10 MPa (Fig. F6). This stress range corresponds approximately to “hydrostatic” depths (the depth if the pore pressure were hydrostatic) of 1 to 1000 m (Fig. F6, right side). The actual effective stress range encountered during Expedition 308 drilling ranges from 0 to ~4 MPa (the difference between the reduced overburden stress and the overpressure at the deepest location at Site U1324) (Fig. F3).

The porosity of the siltstone lithofacies is everywhere less than that of the mudstone or MTD lithofacies (Fig. F6A). However, the siltstone lithofacies declines less with increasing stress than does the mudstone or MTD lithofacies. At a given effective stress, the porosity of the MTD is ~5 porosity units less than the mudstone.

Permeability declines with increasing effective stress for each lithofacies because they have lower porosity at higher stresses (Fig. F6A, F6B). The mudstone and MTD lithofacies have lower permeability than the siltstone lithofacies because they are composed of finer material (Fig. F6B). At 0.1 MPa, mudstone permeability is less than 1 order of magnitude less than siltstone permeability, whereas at an effective stress of 10 MPa, it is almost two orders of magnitude less. The permeability of the mudstone lithofacies declines more with increasing effective stress than that of the siltstone lithofacies. The mudstone and MTD lithofacies compress more with increasing effective stress than the siltstone lithofacies, and this causes the larger drop in permeability with effective stress (Fig. F6A, F6B). The MTD lithofacies has a lower permeability at a given effective stress than the mudstone lithofacies because it is has lower porosity.

Uniaxial compressibility (m_v) is calculated from the void ratio-effective stress relationship (Equation 2). m_v declines by 3 orders of magnitude from 0.01 MPa to 10 MPa (Fig. F6C). This large decrease in compressibility (increase in stiffness) results from the loss of porosity (Fig. F6A) and consequent increase in grain to grain contact. The compressibility of the mudstone lithofacies is the same as that of the MTD lithofacies. The compressibility of the siltstone lithofacies is less than that of the mudstone lithofacies (Fig. F6C).

The coefficient of compressibility (c_v) is calculated from Equation 3 based on the permeability and compressibility described above and assuming a viscosity of 1 × 10^–3 Pa·s. In the mudstone, between 0.1 and 1 MPa, the coefficient of consolidation (c_v) is fairly constant and equal to 1 × 10^–8 m²/s. At higher and lower values of effective stress, c_v is slightly lower (Fig. F6D). c_v is constant because compressibility and permeability decrease at the same rate as effective stress increases (Equation 3; Fig. F6). The MTD coefficient of consolidation is ~5 times lower. This is driven entirely by the lower permeability within the MTD at a given stress. The coefficient of consolidation for siltstone ranges from 3 × 10^–8 m²/s at 0.01 MPa to 5 × 10^–7 m²/s at 10 MPa.

We calculate the dimensionless time factor T_g (Equation 4). At Site U1324, for the mudstone lithofacies we assume a sedimentation rate of 12 mm/y, c_v = 1 × 10^–8 m²/s and a duration (t) of 50,000 y. We find T_g = 23, which results in λ* = 0.75 at the base of the section; even higher values are present in the shallower section. At Site U1322, for the mudstone lithofacies we assume a sedimentation rate of 4 mm/y and find T_g = 2.5. In contrast, the siltstone lithofacies, with c_v = ~2 × 10^–7 m²/s, produces T_g = 1.0 at; this results in no significant overpressure. Similar calculations were presented by Flemings et al. (2008).

In summary, one of the fundamental results of Expedition 308 is that we have shown that rapid sedimentation of the highly compressible mudstones at Ursa Basin results in significant overpressures from the seafloor downward. This result contrasts with the historical view that overpressure is encountered only at greater depths (Dickinson, 1953). The absence of overpressure at shallow depths on the continental shelf inspired theoretical models suggesting that mudstones could only become pressure seals after their porosity was sufficiently reduced by burial (Mello et al., 1994). Expedition 308 results show that outboard of deltaic systems on the continental slope, rapid deposition of fine-grained and compressible sediment creates ideal conditions for overpressure generation.

Several authors modeled the Ursa system in two dimensions (Binh et al., 2010; Dugan and Stigall, 2010; Stigall and Dugan, 2010; Urgeles et al., 2010). They illustrate that flow focusing is present in Ursa Basin. Stigall and Dugan (2010) show that fluids are driven from the zone where the overburden is thick at Site U1324 toward the section where it is thin at Site U1322 along the more permeable beds of the underlying Blue Unit (Fig. F7B). In addition, they illustrate that in the more permeable siltstones present in the bottom half of Site U1324, overpressure is approximately constant (Figs. F3, F7).

A striking result of our study is that submarine landslides all lie within zones of high overpressure (λ* > 0.6) (purple zones in Figs. F3A, F5A). High pore pressure increases the risk of landslides for a given slope (Dugan and Flemings, 2002). For a constant friction angle (φ_f) and sediment with no cohesion, the factor of safety (FS) is proportional to the overpressure ratio (λ*):

where θ is the surface slope. At Ursa Basin, all locations are stable today (FS > 1.0) (Flemings et al., 2008; Sawyer et al., 2009). However, Stigall and Dugan (2010) showed that the deeper zones at Site U1322 were unstable during times of rapid sedimentation soon after deposition of the Blue Unit and that these zones correlate to locations where MTDs are present. Stigall and Dugan (2010) also suggested that the largest MTD (MTD-2 in Fig. F2B) may have resulted from the combination of moderate overpressure from rapid sedimentation coupled with horizontal earthquake acceleration of a magnitude 5 earthquake within 140 km or a magnitude 4 earthquake within 3 km of the Ursa region. There is no record of earthquakes from the time MTD-2 was formed (27–10 ka). However, seismicity from 1973 to 2009 records an earthquake of magnitude 5 within 140 km of Ursa Basin location (Stigall and Dugan, 2010).

The northern continental slope in the Gulf of Mexico has the lowest gradient of the U.S. continental margins, and it has the largest submarine landslides (McAdoo et al., 2000). We interpret that overpressures precondition the slope for failure and that earthquakes may induce the failures and keep the continental slope at a low gradient. Thus, high pore pressure near the seafloor plays a significant role in controlling where submarine landslides occur and controlling the topographic gradient of the continental slope.

Top of page   |    Previous   |    Next