Maximizing Value from available data via Advanced Geostatistical Inversion in the Growler Field

The Growler field produces oil from the middle Birkhead formation. The main production area is a low relief four-way dip closure consisting of channel reservoir with thickness of ~15-20m that has been mapped from the 3D seismic amplitudes and confirmed by wells. Interpretation of the thin and lower quality oil reservoirs in the form of secondary channel and floodplain sandstone deposits from the seismic has not been successful. The inability to discriminate and delineate the geological and/or fluid facies is the main challenge to further explore and develop the field. The challenge is worsened by the uncertainty in the well logs and the poor-quality nature of land seismic data. An advanced pre-stack geostatistical inversion study has been carried out aiming to solve the observed key issues: i) discrimination of different reservoir facies from elastic properties derived from 3D seismic amplitudes; ii) enhancement of the quality of the seismic to resolve the inherent uncertainty associated with the AVO responses; iii) mitigation of the ambiguity of false AVO anomaly due to carbonaceous shale that had led to unsuccessful drilled well. The applied geostatistical inversion study workflow includes iterative seismic petrophysics and rock physics modeling to produce a good quality and consistent set of well logs; robust seismic data conditioning for removal of coherent and incoherent noises, and alignment of seismic events, with the resultant seismic AVO response calibrated with well data; deterministic inversion of conditioned multiple angle stacks and litho-facies estimation using Bayesian inference to provide understanding on the intricacies of the aforesaid challenges before application of geostatistical inversion. Joint facies and elastic properties inversion facilitated by Bayesian-based geostatistical inversion using Multigrid Markov Chain Monte Carlo algorithm has resulted in highly detailed subsurface facies models that show excellent match at most of the 14 blind wells not used in the study.