Log-Based Porosity of ODP Sites on the Cascadia Accretionary Prism

Abstract

Fluid flow is a key to the dynamics of accretionary prisms, and porosity at drill holes can provide valuable clues to that flow in two ways: (1) lateral changes in porosity/depth trends provide estimates of fluid expulsion volumes; and (2) porosity/velocity relationships can be used to determine two-dimensional porosity structure from seismic velocities. We have determined log- based porosities for four holes on the Cascadia accretionary prism, as well as for nearby reference Site 888. Before interpretation of the geophysical logs, substantial reprocessing was needed, particularly for borehole corrections. We used borehole-corrected gamma-ray logs to determine lithology variations (sand vs. silty clay) at each site; at Site 888 the caliper log was also a useful lithology log. Sonic logs were reprocessed to remove noise. A recent algorithm for combining the high accuracy of deep resistivity logs with the high vertical resolution of shallow focused resistivity logs was successfully employed. We then used estimates of thermal gradient and salinity to determine formation factor from resistivity. One-third to two-thirds of the density logs had to be rejected because of poor pad contact. The remaining portions, after conversion to poros- ity based on index-property measurements of grain density, allowed us to determine the coefficients needed for conversion of formation factor logs to porosity. These resistivity-based logs, which we consider to be our best estimates of porosity variation as a function of depth, agree in general with density and neutron logs, but they are more continuous and higher in signal-to-noise ratio. Core porosities are sim- ilar to the porosity logs, but offset slightly owing to rebound or missed fracture porosity. The porosity logs exhibit dramatic intersite variations in both average porosity and consistency with empirical trends. The sites exhibit no systematic lateral changes in porosity; total porosity (both intergranular and fracture porosity) of the studied prism sediments is comparable to porosities of the unconsolidated reference sediments. Of the published relationships between velocity and porosity, only that of Hyndman, Moore, et al. (1993) fits data from our reference sections reasonably well. Only at reference Site 888 did we identify an influence of lithology on porosity/formation- factor and porosity/velocity relationships. Within the prism, cementation accompanied by fracturing dominates lithology influ- ences on both pore and intergranular-contact geometry. As a result, formation factors and velocities are substantially higher than in unconsolidated reference sediments of the same porosity. For the conversion of seismic velocities within accretionary prisms to porosities, the velocity /porosity relationship based on Leg 146 prism sediments is more appropriate than those from undeformed sediments. We caution, however, that observed variability in cementation, and secondarily in compression- induced compaction and fracturing, creates substantial uncertainty in the local velocity/porosity transform for prism sediments.