Integrating The Geophysical Characterization Of Seismic Thin Beds With Stochastic Reservoir Modeling: A Case Study From The Kutei Basin (Offshore Kalimantan, Indonesia)

Abstract

TX 75083-3836, U.S.A., fax +1-972-952-9435 In exploration plays driven by DHI, seismic thin beds are common targets. When resolution capabilities are comparable to isochrones of remarkable elastic contrasts, reflections tend to fall around tuning thickness, where interference phenomena prevent a correct calibration of amplitudes to reservoir properties. If a de-tuning step is not applied in seismic characterization, wrong estimation of reservoir properties and pay thickness may lead to inaccurate volumetrics and misleading geologic models. These issues were faced in the characterization of a Kutei Basin (Offshore Kalimantan, Indonesia) discovery, where gas sands, acoustically marked by lower P-impedance than bounding shales, respond as bright, single-loop reflections (i.e. seismic thin beds). In order to support appraisal and reservoir modeling activities, our seismic characterization approach combined the following steps: a) Statistical tuning charts realization by extracting reservoir isochron and amplitude; b) Investigation of petro-elastic relations through rock physics modeling; c) Definition of deterministic tuning curves by generating a family of pseudo-reservoir responses via forward seismic modeling and d) De-tuned calibration of amplitude -isochron pairs to Porosity*Thickness (PT) and Net Pay. Appraisal wells confirmed the reliability of this approach but also revealed how other uncertainties may impact reservoir quality predictions. For this reason, PT realizations were only used as "soft" information in the reservoir model, built through a Sequential Indicator Simulation of sedimentological facies recognized in cores. Assuming a link between seismic porosity and facies, normalized porosity maps derived from PT estimates were used to condition the horizontal facies distributions and define a seismic-consistent variography. Sensitivity analysis confirmed the robustness of the approach and final results appear consistent with the de-tuned PT predictions. In a field where seismic thin beds affected by tuning represent the key geophysical issue to address, the applied workflow ensured the link between geophysics and geological modeling, from assessment of gas in-place to dynamic simulation.