Arthit Log-Derived Permeability Modeling Project

Abstract A log-derived permeability prediction model for the new Arthit Field in the Gulf of Thailand has been successfully developed based on available cores and well test results. Accounting for more rock and fluid parameters than just porosity alone, the log-based algorithm has proven to be better-equipped to predict permeability with less uncertainty than traditional K-PHI cross-plots. The Arthit permeability model considers not only porosity, but variations in clay mineral composition and clay volume. It is a well-known fact that sands having the same porosity will not necessarily have the same permeability due to other reservoir factors affecting permeability. A traditional K-PHI cross-plot employing a straight-line correlation of permeability and porosity cannot be expected to provide the best estimate of rock permeability due its indifference to changes in the rock matrix. The Arthit log-derived permeability model (K-Log) considers not only porosity, but variations in clay mineral composition and clay volume. It eliminates the need to subdivide data by rock type, does not require a semi-log scale correlation, and reduces the uncertainty of prediction. Clay diagenesis and reservoir temperature were identified as important parameters affecting permeability where predictions in shallow, low-temperature zones poorly matched actual measurements but reliably matched measurements at higher temperatures. The model currently contains vertical definition by geologic formation. Areal definition should be possible in the near future as more permeability measurements are gathered from various geologic areas within Arthit resulting in distinct mineral factor sets for major geologic areas such as graben, basement flank, or basement high. Engineering applications currently planned for permeability modeling include well completion design, high-CO2 gas blending, field production forecasting, sand control, and reservoir simulation.