Reservoir Characterization through Bayesian Joint Inversion of Seismic and EM Attributes

Abstract

Abstract The main objective of the reservoir characterization is to provide reservoir property maps to reservoir engineers for the field appraisal. Since that porosity and water saturation affect both seismic and EM attributes, their estimation can reduce the uncertainty of the interpretation and consequently the costly drilling of un-productive reservoirs. This study explains the petrophysical joint inversion (PJI) of seismic and EM attributes to estimate the petrophysical model in terms of porosity and water saturation. Rock physics models are involved as forward models to form a proper link between data input (seismic and EM attributes) and the petrophysical parameters, (porosity and water saturation). According to the formal Bayesian theory, the uncertainty affecting the input data (AI and resistivity models) and the model parameters (porosity and water saturation) is captured by, respectively, data and model covariance matrix. It is assumed Gaussian probability density functions for model parameters and input data. PJI is applied where both seismic attributes and resistivity models show consistent anomalies as evidence of a potential reservoir. PJI objective is to jointly estimate porosity and water saturation from AI (seismic inversion derived), and resistivity (CSEM inversion derived), by using rock physics modeling within Bayesian framework. The result is represented by the porosity and water saturation distribution within the potential reservoir. PJI realizes a multi-physic inversion able to exploit the complimentary information available in the single physic domains, which are seismic and EM. The methodology is applied to a real hydrocarbon exploration scenario to evaluate its contribution to the interpretation phase. 3D volumes of estimated porosity and saturation show how the joint inversion of acoustic impedance and electrical resistivity provide a quantitative description of the reservoir properties and with it a measure of uncertainty, which is consistent with the petrophysical model and observations. Introduction Reservoir characterization objectives are to estimate the petrophysical properties of the prospective hydrocarbon traps and to reduce the uncertainty of the interpretation. In this framework, we present a workflow for petrophysical joint inversion of seismic and EM attributes to estimate the petrophysical model in terms of porosity and water saturation. This study realizes the joint inversion within the probabilistic structure provided by the Bayesian theory. 3D volumes of estimated porosity and saturation, show how the joint inversion of acoustic impedance and electrical resistivity can provide a quantitative description of the reservoir properties and with it a measure of uncertainty, which is consistent with the petrophysical model and observations.