Deterministic and stochastic seismic inversion methods for gas discrimination at La Creciente Field, Colombia

Abstract

Summary We applied deterministic and stochastic seismic inversion methods to 3D data of La Creciente field in Colombia, with focus on gas and lithology identification. Our deterministic inversion considers an elastic isotropic medium model to jointly estimate Vp, Vs, density and source wavelet deviations. We estimate the probability of the gas sands with multivariate Gaussian classification based on seismicderived elastic properties and petrophysical characterization. Our study shows that Cienaga de Oro gas reservoir is characterized by lower density, moderately lower Vp and moderately larger Vs than surrounding strata and non-producing Cienaga de Oro intervals. To further exploit the seismic and petrophysical information, we applied a stochastic petrophysical seismic inversion. The method combines the geostatistical modeling of reservoir parameters (porosity, shale fraction and saturation), a rock physics model for the prediction of the elastic properties and seismic modeling. Using a Markov Chain Monte Carlo method we generate a large number of realizations of the reservoir and elastic parameters that jointly honor the seismic observations, the rock physics model and the prior information. With this approach we describe the posterior uncertainties of the reservoir properties including the gas saturation.