Back to Table of Contents

44476-AC8
Overpressure and Slope Stability in the Deepwater Gulf of Mexico

Peter B. Flemings, University of Texas (Austin)

High overpressure near the seafloor where multiple submarine landslides have occurred on the continental slope, offshore Louisiana, Gulf of Mexico

This project focused on developing an understanding of the linkage between sedimentation, overpressure and slope stability. Our contributions are four-fold: 1) measurement of in-situ pore pressure; 2) characterization of submarine landslides; 3) measurement of geotechnical properties; and 3) forward modeling to link laboratory analysis and field observations.

1)              Pore Pressures in the Ursa Basin: Overpressures measured with pore pressure penetrometers during Integrated Ocean Drilling Program (IODP) Expedition 308 reach 70 % and 60% of the hydrostatic effective stress  in the first 200 meters below sea floor (mbsf)  at Sites U1322 and U1324, respectively, in the deepwater Gulf of Mexico, offshore Louisiana.  High overpressures are present within low permeability mudstones where there have been multiple, very large, submarine landslides during the Pleistocene. Beneath 200 mbsf at Site U1324, pore pressures drop significantly: there are no submarine landslides in this mixture of mudstone, siltstone and sandstone. We interpret that the high overpressures observed are driven by rapid sedimentation of low permeability material from the ancestral Mississippi River.

2)              Submarine Landslides in the Ursa Basin: In the Ursa Basin (Gulf of Mexico), we interpret that Mass Transport Deposits (MTDs) record failures that mobilized along extensional failure planes and transformed into fluidized, long-runout, flows. Failure proceeded retrogressively: scarp formation subjected adjacent sediment to undrained unloading (Rankine active failure), which drove successive scarp formation updip. We used 3-D seismic reflection data, core and log data from IODP Expedition 308, and triaxial shear experiments to develop this retrogressive model. MTDs are imaged seismically as low-amplitude zones above continuous, grooved, high-amplitude basal reflections. Two seismic facies characterize the internal structure: 1) a Chaotic facies, typical of the downdip region and 2) a Discontinuous Stratified facies, typical of headwall/sidewall regions. The Chaotic facies contains discontinuous, high-amplitude reflections that correspond to sinuous, flow-like features in interval amplitude maps. This facies has higher bulk density, shear strength, and resistivity than bounding normally consolidated sediment. In contrast, the Discontinuous Stratified facies has only slightly higher bulk density and resistivity than the bounding sediment. This facies contains relatively dim reflections that abut against in-tact blocks of stratified reflections. Deformation is limited to slightly tilted bedding and small-offsetfaults. In both facies, densification is greatest at the base, resulting in a strong basal reflection. Undrained triaxial shear tests document strain weakening (sensitivity). For an undrained failure at 30 meters below seafloor, (overpressure = 70% of hydrostatic effective stress and sensitivity = 3), soil would weaken to a condition in which the gravitational shear stress exceeds soil strength. In this condition, soil is susceptible to flow.

3)              Rock Properties in the Ursa Basin: We also conducted extensive uniaxial consolidation tests on whole core samples to obtain the consolidation properties of the Ursa mudstones. The results suggest that the compression index linearly decreases with in situ void ratio. This implies that a locally-defined virgin compression curve cannot validly be extrapolated over a large range in effective stress. This effect is particularly important at shallow depth where void ratio decreases rapidly. We have shown that the relationship of compressibility index versus void ratio can be obtained from a single consolidation test by compressing the soil over a large range in effective stress. A virgin compression curve can then be constructed based on this relationship to predict pore fluid pressure. In the Ursa Basin, this new approach successfully predicted pressures interpreted from the penetrometer measurements within the non-deformed sediments. The mass transport deposits appear to be more compacted than the non-deformed sediments. The virgin compression curve based on the assumption of uniaxial strain underpredicts the in situ pressure in the mass transport deposits.

4)              Modeling Ursa Basin Pore Pressures: The average sedimentation rate from the seafloor to the top of the Blue Unit is 12 mm/year at Site U1324 and 3.6 mm/yr at Site U1322. Ursa mudstones have hydraulic diffusivities of 2 x 10^-8 m²/s whereas siltstones have diffusivities greater than 2 x 10^-7 m²/s. We interpret that at Site U1324, rapid deposition of fine grained low permeability mudstone in the upper 200 mbsf generated the high overpressures present. Reduced overpressure at depth at Site 1324 suggests suggest lateral flow (drainage) whereas high overpressure at Site 1322, where the sedimentation rate is low requires inflow from below: lateral flow in the underlying permeable aquifer provides one mechanism for these observations. High overpressure near the seafloor reduces slope stability and provides a mechanism for the large submarine landslides and low regional gradient (2 degrees) offshore from the Mississippi delta.

Back to top