Abstract

Real-time mapping of sedimentary formations, detection of structural anomalies and evaluation of formation properties can be achieved through logging while drilling (LWD) resistivity inversion. We present the physics-based framework that relies on the predictor-corrector methodology, sliding windows of tool signals and formation parameters and bonded by regularized optimization. Further, we outline the underpinning methods constituted in accelerated forward solver and inversion with focusing (narrowing) of solution equivalency. The steering and formation evaluation results are explained with the use of the classic Oklahoma formation model and field-measured data from the Sabah Basin in Indonesia. Multi-resolution workflows employing the assemblies of shallow-to-extra deep resistivity logging tools are examined. The proposed approach suggests an effective means for look-around and prediction-ahead LWD imaging.

Introduction

Reservoir navigation, in a hydrocarbon reservoir, relies on the relationship of the LWD tool responses and the parameters of the surrounding formations. The complexity and challenges of these reservoirs are gradually increasing in exploration and exploitation both technically and environmentally (Horstmann et al., 2020). While the majority of hydrocarbon bearing deposits have a layered structure of varying layer properties in 3-D space, so-called 3-D quasi-layered formations, the anomalies such as intermittent shale-sand interfaces, fault zones, pinch outs, embedded stringers and salt domes, are becoming increasingly common in more challenging reservoirs (Wilson, et al., 2019). A more accurate, multi-dimensional, representation, with the use of modern visualization software and computationally accelerated inversion with robust look-around and prediction-ahead capabilities becomes a demanding proposition for confident well placement (Maraj et al., 2021).

The ultra-deep azimuthal/directional resistivity (UDAR/EDAR) technology has been in use for 10+ years. The conventional hardware design employs induction transmitting and receiving antennas. Such acquisition systems are relatively insensitive with respect to fluctuations in excitation and registration conditions and supports the wellbore and mud compensation thru coil superposition (Hartmann et al., 2014). The concept of signal registration with electric dipole antennas has also been investigated, where the prediction-ahead sensitivity with respect to a single boundary is demonstrated with the numerical simulations (Fan et al., 2021).

This content is only available via PDF.
You can access this article if you purchase or spend a download.