Key points are not available for this paper at this time.
Abstract Low-permeability laminated reservoirs have been a challenge for accurate characterization, more so when realtime decisions are critical for operations. Innovative and automated logging-while-drilling (LWD) measurements have come a long way to address this challenge. This work presents a study from the Alaska North Slope where two sections of a well were drilled with nuclear magnetic resonance (NMR) and bore-hole images (BHI) in conjunction with triple combo well-logs to characterize the challenging reservoirs. In this study, the traditional measurements were supplemented with the new generation nine wait-time LWD NMR measurement as well as high-resolution LWD images. Realtime streaming of both NMR and image logs enabled by innovative data compression methodology resulted in automated answers as the intervals in the subsurface were getting drilled. Realtime porosity-distributions were made available with associated answer products of bound and free fluid volume and permeability of the thin sands, validated with realtime borehole images. The data were processed using a rate of penetration (ROP) aware 2D inversion (T1 and T2) to correct for any logging speed effects. Subsequently, to fully unlock the potential of the acquired nine wait-time NMR data in conjunction with the other measurements, special methodologies were employed including "Factor Analysis" using a parallel analysis-based application and Viscosity Transforms based on empirical relations. These techniques enabled a comprehensive analysis and revealed a degree of variability in fluid viscosity along the drilled section, hitherto not understood. This variability is considered crucial for making informed completion decisions. It is important to note that such analysis would not have been possible earlier during drilling, enabling realtime decisions in thin beds. Integration of new technology LWD NMR with borehole images revealed critical insights for development of low-permeability laminated reservoir, validated with core data.
Akindipe et al. (Mon,) studied this question.