Probabilistic facies analysis using 3D crossplot of stochastic forward modeling results

Abstract

The facies analysis using stochastic forward modeling is an effective method when well data are too sparse. Using this method, we can analyze the distribution of facies probabilistically. However, when the result of stochastic forward modeling is plotted on the 2D crossplot, the boundaries of the facies probability density functions can be overlapped each other and ambiguous to distinguish. To overcome this problem, we propose the probabilistic facies analysis method using the 3D crossplot of stochastic forward modeling results. In the proposed method, three axes of 3D crossplot consist of the parameters with which the facies is well separated on the 3D crossplot. In this study, acoustic impedance (Ip), pseudo gamma ray (GR) log, and pseudo water saturation (Sw) log axes were used for the 3D crossplot. To perform stochastic forward modeling, pseudo GR log and pseudo Sw log must be expressed mathematically with seismic attributes. We use linear multi-regression analysis of well logging data to derive the mathematical relationship between them. Additionally, to estimate the facies using seismic data, we extracted pseudo GR and pseudo Sw log data from seismic data using probabilistic neural network (PNN) prediction. Finally, from the results of applying the proposed facies analysis method to the field data, we can confirm that the proposed method is more effective than the conventional facies analysis method using the 2D crossplot of stochastic forward modeling results.