´, Pc´, and the OCR, as well as the sediment compressibility coefficients (Cc and Cr) and the calculated eo.
Consolidation tests were performed on 12 samples from five of the major lithostratigraphic units used to describe sediments recovered during Leg 207. The deepest sample came from Site 1261 at a depth of 364 meters composite depth (mcd) and the shallowest from 26 mcd at Site 1259. A summary of consolidation results is presented in Table T1 and includes the stress history–related calculations of in situ ´, Pc´, and the OCR, as well as the sediment compressibility coefficients (Cc and Cr) and the calculated eo.
No corrections were applied for the effects of sample disturbance. Disturbances related to coring, sampling, and preparation of the sample all result in an underestimate of Pc´ and reduce both eo and Cc. Therefore, disturbance masks the effects of overconsolidation in test specimens. A qualitative approach for evaluating sample disturbance is assessing the shape of the consolidation curve (e-log[´]). A sharp transition into the virgin compression curve suggests a high-quality sample. A more quantitative approach is by calculating the amount of volumetric strain that accumulates prior to reaching Pc´ (Lunne et al., 1999). Developed primarily for the evaluation of samples from near-surface sediments and soils, Lunne's method evaluates sample quality using ratings between excellent and very poor. Evaluation includes consideration of the OCR, with higher levels of strain permitted for samples with low OCR values (Table T2). The majority of the samples tested here experienced between 5% and 10% strain before reaching Pc´. Using the evaluation criteria of Lunne et al. (1999), the quality of these samples range from poor to good/fair (Table T1). The relatively high degree of sample disturbance results from the depth of burial of the samples and the type of sampling tools used to collect them, in this case the rotary core barrel system. In the majority of marine sediments, there exists an order of magnitude difference in Cr and Cc (Holtz and Kovacs, 1981). At Site 1261, the high elasticity of the Neogene nannofossil clays reduced the difference between Cr and Cc, thus increasing the level of apparent sample disturbance.
Significant levels of overconsolidation, as evidenced by the calculated OCR (see Table T1), occur in samples from Sections 207-1257C-1R-2, 207-1257C-6R-3, and 207-1259A-3R-5. These results are consistent with an interpretation of mass wasting on the eastern flank of the rise based on seismic reflection data (Fig. F5) (Shipboard Scientific Party, 2004b).
An estimate of the magnitude of the erosional events is calculated by determining the amount of material required to account for the difference between Pc´ and a hydrostatically determined ´ (Table T1). The removed overburden is added to the current effective stress profile to calculate the corrected effective stress profile. Predictions of ein situ are made by combining the corrected effective stress profile with the compression indexes derived from consolidation tests using:
´), and (17)
´), (18)
where
A close agreement between the predicted and measured e supports the stress history interpretation and permits the application of the test-derived e-log(k) relationships to determine the current in situ k profile.
Permeability results are summarized in Table T3 and include calculated variables (Cv and Av) used to derive k from the time deformation data, as well as the calculated k for each set of low-gradient flow pump tests. A table is also presented summarizing q, i, and k for each set of flow pump tests (Table T4). Individual sample results from consolidation and hydraulic conductivity tests are presented in Figure F6.
One sample, Sample 207-1261A-4R-2, 80 cm, was tested using both the lever arm consolidometer and the Geocomp system. Direct measurements of k were performed on the lever arm system during test B. Prior to flow measurements, the sample was unloaded to 30 kPa, when a leak was detected in the consolidation cell. After the sample was reloaded and consolidation began to proceed along the virgin compression curve, k tests were started.
Very low permeability in a few samples either reduced the number of flow rates applied at any given e or, in one case, precluded the use of low-gradient flow pump measurements. A steady-state hydraulic gradient could not be established across Sample 207-1258B-36R-2, 8 cm (344.49 mcd), a Maastrichtian nannofossil chalk with forams. Flow pump tests were not possible on two other samples, Samples 207-1261A-13R-4, 140 cm, and 20R-1, 140 cm, because of leaks that developed in the consolidation cell. However, results from samples taken from this same lithostratigraphic unit at Site 1261 show excellent agreement between the derived and measured permeability determinations and suggest that the derived k values are correct.
The greatest offsets between the derived and measured k values occurred in Samples 207-1257C-1R-2, 140 cm, and 13R-1, 130 cm. In both cases, the flow pump measurements were selected as being representative of the sample's permeability. The ability to derive k from Terzaghi's one-dimensional theory of consolidation requires that over the analyzed period of settlement, changes in void ratio maintain a unique relationship with changes in effective stress (Holtz and Kovacs, 1981). If secondary consolidation occurs (i.e., the rearrangement of particles within the sediment), both the void ratio and permeability can change without accompanying changes in the effective stress and this assumption is no longer valid (Holtz and Kovacs, 1981). In marine sediments, permeability estimates made using Terzaghi's theory tend to be slightly lower than the actual permeability (by a factor of 2–4), with secondary consolidation consistently playing a small role in the deformation of a sample under a load (MacKillop, 1995). In organic-rich soils, secondary compression can be greater than primary consolidation during loading (Hobbs, 1986). Sample 207-1257C-13R-1, 130 cm, was a black shale with a TOC content of 8.2 wt%. Differences between the flow pump measured and derived permeability curves for this sample may be associated with a high level of secondary compression during loading. Although secondary consolidation may have been a factor in the underestimated k for Sample 207-1257C-1R-2, 140 cm, it should not have been as significant as in the high organic content black shale sample. Other assumptions of Terzaghi's theory that may not have been valid during the testing of Sample 207-1257C-1R-2, 140 cm, include: (1) that the sample be homogeneous and 100% saturated, (2) that the solids are incompressible, and (3) that drainage occurs from compression in one dimension (Terzaghi, 1943).
The other sample from the black shale sequences was Sample 207-1258B-45R-4, 45 cm (TOC = 6.3 wt%). During testing, a leak was discovered around the perimeter of the sample. The leak prevented the establishment of a hydraulic gradient across the sample. After unloading and sealing the leak, hydraulic conductivity measurements were conducted at 6144, 7140, and 8100 kPa. Results from flow pump tests on Sample 207-1257C-13R-1, 130 cm, are used to construct e-log(k) relationships for the black shales. Estimates from these tests give a lower hydraulic conductivity for void ratios greater than ~0.7 and higher estimates when the void ratio is less than ~0.7 when compared to the results from Sample 207-1258B-45R-4, 45 cm (Fig. F4). The selection of this sample as being representative of the black shales at Sites 1257, 1259, and 1261 reduced the variability in the estimates of permeability within and between sites, perhaps masking the potential for high and low permeability lenses.
Lithostratigraphic Unit III sediments on Demerara Rise generally consisted of Maastrichtian–late Paleocene nannofossil chalks and clays. These sediments exhibited the lowest porosities recovered on the rise and often showed decimeter-scale variations between chalks and clays (Shipboard Scientific Party, 2004c). Two samples were taken from this unit, Samples 207-1257C-6R-3, 140 cm, and 207-1258B-36R-2, 45 cm. Both samples exhibited significantly different compressibility and permeability behavior, with the sample from Site 1258 being less compressible and less permeable, a state that may be linked to the diagenetic history of the sample. In constructing e and k profiles for the Unit III sediments, Cc and e-log(k) from Sample 207-1257C-6R-3, 140 cm (Tables T1, T3), was used at Sites 1257, 1259, and 1261. However, e-log(k) relationships from Sample 207-1258B-36R-2, 45 cm, were also plotted to give a lower limit to the permeability of sediments across this unit.
