Applicability of experimental techniques is critical to accuracy of results. In this paper, a comparison was made between nuclear magnetic resonance and mercury injection porosimetry to clarify their applicability in pore structure characterization of tight oil reservoirs. Five tight sandstone core samples from the Upper Triassic Ordos Basin in China were collected for nuclear magnetic resonance, pressure-controlled porosimetry and rate controlled porosimetry. The T2 relaxation time distribution of nuclear magnetic resonance was compared to the pore size distribution calculated by a combination of pressure-controlled porosimetry and rate-controlled porosimetry.

The T2 distribution of sample Z211-1, N78-3 and T2-1 show three peaks, i.e. ~ 1 ms, ~100 ms and ~1000 ms. However, the T2 distribution of sample Z147-2 shows unimodal and that of sample Z117-3 shows bimodal, and the amplitudes are much smaller. For pressure-controlled porosimetry, the average final intrusion mercury saturation exceeds 80%, while the extrusion efficiency is low, indicating large pore-throat radius discrepancy. For rate-controlled porosimetry, the final intrusion mercury saturation averages 48.20%, and most of the mercury enters throats. The nuclear magnetic resonance T2 distribution and pore size distribution by a combination of pressure-controlled porosimetry and rate-controlled porosimetry show reasonable agreements. Rate-controlled porosimetry tends to overestimate the radius of pores, while some clay dominated pores are not water saturated. These might be the causes of differences between the nuclear magnetic resonance T2 distribution and the mercury injection porosimetry pore size distribution.

It is concluded that a combination of PCP and RCP can provide more accurate pore size distribution of tight sandstone, while nuclear magnetic resonance seems to be a fast and economical method.

You can access this article if you purchase or spend a download.