Permeabilities are reported at effective pressures of 15 MPa. For the majority of the samples, pressure had little effect on the results of the measurements (Fig. F2). For these samples, permeability was relatively constant over the span of the test, both with pressure increases and decreases. Several samples did respond to pressure changes, with permeability decreasing as pressure increased (Fig. F3). For most of this subset of samples, the permeability change was <50%; however, for five samples the change in permeability was close to or greater than an order of magnitude (Samples 193-1188A-3R-1, 13–15 cm; 193-1189A-12R-1, 35–37 cm; 193-1188A-14R-1, 102–104 cm; 193-1188F-3Z-2, 121–123 cm; and 193-1189B-13R-1, 54–56 cm). The decrease in permeability is likely caused by a reduction of interconnected pore space caused by the increase in pressure. The latter three samples were damaged during the testing procedure, which would contribute to the decrease in measured permeability. The variation in the response to pressure change between the samples may be dependent on the lithology, amount, or style of alteration; the amount of interconnected porosity; or a combination or factors. It may be noted that all of the complete measurements from Hole 1189A decreased with increasing pressure. However, there is no petrological data from the samples to determine the cause of the differences. Many of the samples affected by pressure changes did not regain permeability as pressure was decreased (i.e., Samples 193-1188A-21R-1, 82–84 cm). This may be because of structural degradation induced by the high pressures used in the measurement process.
Six samples were damaged during testing due to collapse or compaction that caused a permanent reduction in permeability before a reliable measurement could be made. Although we are not confident in the measurements for these samples, they are included in the following graphs and marked as damaged. Values for these samples are not included in the discussion of the results. Damage to the samples occurred during permeability testing, which was made after the porosity calculations; therefore, all porosity values are representative of the samples. Regardless, samples are marked as damaged in both porosity and permeability plots for consistency.
Both permeability and porosity measurements of the samples were highly variable between samples. Permeabilities measured in reliable tests ranged from ~1.4 x 10–19 to 7.0 x 10–15 m2 (Table T1), with an average of 4.5 x 10–16 m2. Broadly, permeability decreases with depth, but with considerable scatter of the core-scale measurements (Fig. F4). Permeability values are more tightly grouped near the seafloor and become more scattered as depth increases. Correlations to other physical properties, such as thermal conductivity, velocity, vesicularity, composition, or grain density, are weak if present. Qualitative examinations of thin sections from a subset of the minicore samples (samples with compositional analyses in Table T1) indicate that microscopic features such as crystal size and structural fabric also do not have obvious correlations to the permeability.
Porosity values calculated at NER vary from ~1% to 43% (Fig. F5), with an average of 21%. Values calculated at NER are similar to shipboard measurements, which ranged from <1% to 47% for the same sample set. Permeability decreases with decreasing porosity (Fig. F5), although at midrange porosities and permeabilities the values are more scattered. The one low porosity, high permeability sample is from ~372 mbsf and has 26% relict plagioclase based on thin section analysis, indicating that it is much less altered than other samples of similar depth.