Previous   |    Next

doi:10.2204/iodp.proc.322.208.2013

Laboratory testing methodology

Whole-round core samples from Sites C0011 and C0012 (Fig. F1; Table T1) were capped and sealed after collection and stored at 4°C to help maintain the natural water content. All whole-round samples were selected to isolate samples that were hemipelagic mudstone. Samples were taken out of the sealed core liner to conduct CRS consolidation experiments and flow-through permeability experiments following American Society for Testing and Materials (ASTM) International standards (ASTM International 2004, 2006). All experiments are performed at room temperature (20°C).

CRS consolidation experiments

The consolidation system at Rice University (USA) can achieve a maximum consolidation stress of ~4200 kPa, which equates to a burial depth of ~500 mbsf for normal hydrostatic conditions. We limit CRS experiments to specimens from depths shallower than 300 mbsf to allow for elastoplastic consolidation even if specimens are overconsolidated. CRS experiments were completed in a rigid confining ring to maintain the uniaxial strain of the specimen. Each specimen was trimmed using a trimming jig, a wire saw, and a sharp-edged spatula to minimize disturbance during preparation and to provide a specimen diameter that was the exact diameter of the rigid confining ring. Samples were trimmed in a vertical orientation to allow assessment of vertical permeability (k_v) or in a horizontal orientation to allow assessment of horizontal permeability (k_h). Once the specimen was in the confining ring, a wire saw, a sharp-edged spatula, and a recess tool were used to make the specimen into a right cylinder of a fixed height. The use of the trimming jig, confining ring, and recess tool facilitate making specimens of identical diameter and height. Each specimen had an initial height (H_o) of 2.41 cm and an initial diameter of 5.09 cm. A constant, controlled cell pressure (P_c) (386 kPa) was applied to each specimen to ensure saturation with an idealized seawater solution. An initial saturation period of at least 8 h was used to make sure the specimen was at 100% saturation. Then the specimen was uniaxially deformed at a constant rate of strain. The strain rate () was adjusted for each specimen to ensure a pore pressure ratio <0.10 (ASTM International, 2006). The pore pressure ratio depends on the strain rate and the permeability of the specimen. Total axial stress (σ_a), instantaneous sample height (H), and basal pore pressure (P_p) were recorded throughout the experiment. Each experiment was completed at a maximum consolidation stress exceeding the hydrostatic effective vertical stress (σ_vh′) for the specimen; σ_vh′ is total vertical stress less hydrostatic fluid pressure and is determined by integrating the bulk density data. The maximum consolidation stress (σ_v′_max) for each experiment was chosen to be at least three times σ_vh′ for the specimen in that experiment (Table T2). This allowed reaching elastoplastic or virgin consolidation even for samples with mild overconsolidation. Routine inspection of consolidation curves confirmed that each specimen reached elastoplastic deformation. Total vertical stress was determined from bulk density (ρ_b) data (see the “Site C0011” and “Site C0012” chapters [Expedition 322 Scientists, 2010a, 2010b]). Hydrostatic fluid pressure was calculated assuming a constant seawater density (ρ_w = 1024 kg/m³).

CRS consolidation experiments provide data to constrain hydraulic conductivity (K) for laboratory conditions (ASTM International, 2006), which can be converted to permeability (k = Kµ/ρ_wg) based on the water density (ρ_w = 1024 kg/m³) and viscosity (µ = 0.001 Pa·s) at laboratory conditions:

k = HHoµ/2Δu,

(1)

where

• = strain rate,

• H_o = initial specimen height,

• H = instantaneous specimen height, and

• Δu = base excess pressure.

Base excess pore pressure is defined as the difference between basal pore pressure and cell pressure (P_c):

A smoothed base excess pore pressure, based on a three-point moving average, is used. A six-point moving average is used to smooth the strain rate.

We use permeability-porosity (k-ϕ) data during normal consolidation to define a log-linear relation between k and ϕ for each specimen (e.g., Neuzil, 1994; Schneider et al., 2011). Each specimen-specific model is used to estimate permeability at the in situ porosity. We assume that the porosity at the laboratory-determined preconsolidation stress of each specimen represents the in situ porosity. Preconsolidation stresses were determined using the work-stress method (Becker et al., 1987). This method defines the preconsolidation stress as the intersection of the linear extensions of the elastic and elastoplastic deformation curves in work-stress space, which provides a well-defined estimate the preconsolidation stress even if the transition from elastic to elastoplastic deformation is not well defined. Laboratory-derived in situ porosity for each specimen compares well to shipboard porosity data (Fig. F2). Initial specimen porosity was determined in our laboratory from mass and density measurements following the approach presented by Blum (1997). As the sediments are clay rich containing smectite and illite, these shipboard and shore-based porosity data may overestimate the in situ porosity. The exact magnitude of porosity change is controlled by the sample porosity and smectite content. Previous studies of Nankai sediment show that corrections for smectite content can yield porosity up to 16% lower than shipboard-determined porosity (e.g., shipboard porosity of 33% and smectite content–corrected porosity of 17%) (Gamage et al., 2011). Porosity corrections for the samples in this study could be made based on the bulk sediment smectite content, which is 33–55 wt% for samples from Site C0011 and 31–55 wt% for samples from Site C0012 (Underwood and Guo, 2013). In addition, shipboard mass and density porosity data have significant variability that has been interpreted as a result of significant coring disturbance (see the “Site C0011” and “Site C0012” chapters [Expedition 322 Scientists, 2010a, 2010b]). This suggests that mass and density and laboratory porosity are a high estimate of the in situ porosity. Therefore, when we refer to in situ permeability, it should be noted that it is the estimated permeability at the high-end estimate of in situ porosity for all specimens. Porosity during consolidation was determined using the strain data. This approach allows us to present the horizontal or vertical permeability for each CRS specimen, depending on its orientation, at the in situ porosity and at its sample depth (Table T2; Figs. F3, F4). For depths where a horizontal and a vertical specimen were tested, we also provide the ratio of these permeabilities (k_v/k_h) (Table T2).

Flow-through permeability experiments

Permeability experiments are completed on cylindrical specimens in a flexible-wall membrane. Each specimen was trimmed using a trimming jig, a wire saw, and a sharp-edged spatula to minimize disturbance during preparation and to provide a uniform specimen diameter. Samples were trimmed in a vertical orientation to allow assessment of k_v or in a horizontal orientation to allow assessment of k_h. Specimen length and diameter varied depending on the ability to get an intact specimen (influenced by sample disturbance from drilling and coring), but in general, diameter was near 5.08 cm and length was at least 3.86 cm. Each specimen was placed in an impermeable, flexible-wall membrane, and porous stones were placed on the ends. Each specimen was then loaded in the permeability chamber where it was connected to three pumps. One pump controlled the pressure at the top of the specimen, one pump controlled the pressure at the bottom of the specimen, and one pump controlled the isostatic confining pressure around the specimen. Deionized water was used as the confining fluid, and an idealized seawater solution was used as the permeant. The confining chamber was filled, and constant pressures were applied to the top and bottom of the specimen (276 kPa) and to the confining fluid (310 kPa) to maintain a constant isostatic effective stress (34 kPa) while maintaining 100% water saturation on the specimen. After a minimum of 8 h for the saturation phase, each specimen was isostatically consolidated to an isostatic effective stress of 276 kPa by increasing the pressure of the confining fluid and maintaining constant pressure in the sample (i.e., top and bottom pumps). This consolidation phase was applied to close any microcracks in the sample that may have been created by the coring or recovery processes, such that during the permeation phase, we observed the in situ permeability. As our isostatic effective stress is lower than hydrostatic effective stresses, some microcracks may not have closed during consolidation. After stabilization at an isostatic effective stress, flow-through permeability tests were conducted by fixing the pressure gradient across the specimen while maintaining the isostatic effective stress at the center of the specimen. The inlet and outlet fluid volumes were monitored during the experiment, and all experiments were run until steady-state was achieved (inflow and outflow volumes were equal while the sample length and pressure difference remained constant). From the steady-state flow rate, known sample dimensions, and fixed pressure difference across the specimen, the permeability of any specimen was calculated by rearranging Darcy’s law:

where

• Q = steady-state flow rate,

• µ = fluid viscosity,

• A = area of the specimen,

• Δl = length of the specimen, and

• ΔP = fixed pressure difference across the specimen.

Fluid viscosity (µ) was adjusted for temperature, which was measured continuously during the experiments. Viscosity values did not statistically diverge from a value of 0.001 Pa·s (see PERM in “Supplementary material”), which reflects standard laboratory conditions. Sample length was also recorded during the experiments. Sample area was assumed to be constant from the start of the experiment, and was determined from the average of five measurements of sample diameter.

Each calculated permeability is presented at the specimen porosity and depth (Table T3). Specimen porosity after the flow-through experiment was determined in our laboratory from mass and density measurements following the approach presented by Blum (1997). All mass and density calculations assume standard seawater density (1024 kg/m³) because samples were saturated with an idealized seawater solution. Interpreted laboratory porosity is similar to the porosity of nearby samples analyzed during Expedition 322 (Fig. F2). Specifically, the laboratory-determined mass and density porosity values are in the range of the shipboard-determined mass and density porosity. It is important to note, however, that the shipboard mass and density porosity data have a significant amount of scatter that has been interpreted as a result of significant coring disturbance (see the “Site C0011” and “Site C0012” chapters [Expedition 322 Scientists, 2010a, 2010b]). Based on the similarity of shipboard and laboratory porosity data, we infer that the laboratory samples also suffer from moderate to significant sample disturbance. This sample disturbance yields a mass and density–based porosity that is likely a high-end estimate of the in situ porosity. Bulk smectite content of samples ranged from 33 to 55 wt% for samples from Site C0011 and from 31 to 55 wt% for samples from Site C0012 (Underwood and Guo, 2013). No porosity correction was made for smectite content, so laboratory-based porosity will overestimate true sample porosity (Gamage et al., 2011). Therefore, all references to in situ porosity reflect a high-end value for the in situ porosity.

Top of page   |    Previous   |    Next