Prestack seismic joint inversion of reservoir elastic and petrophysical parameters using deterministic optimization method
Li Zhi-Yong,Qian Feng,Hu Guang-Min,He Zhen-Hua
DOI: https://doi.org/10.6038/cjg20150521
2015-01-01
Abstract:Elastic and petrophysical parameters of reservoir can be directly applied to lithology prediction and fluid identification, they are the basic elements in both comprehensive evaluation and fine characterization of hydrocarbon reservoir. Existing seismic joint inversion methods for estimating elastic and petrophysical parameters of reservoir are mainly based on either Gassmann equation, with which these parameters are inversed from prestack seismic data by stochastic optimization methods; or Wyllie equation, with which these parameters are inverted from poststack seismic data by deterministic optimization methods. The purpose of this work is to develop a strategy for estimating elastic and petrophysical parameters of reservoir based on Gassmann equation with deterministic seismic prestack inversion. We use Gassmann equation to build up the relationship between prestack seismic data and petrophysical parameters. We treat the joint posterior probability of elastic and petrophysical parameters as the objective function under the Bayesian architecture, by expanding the the objective function with Taylor formula, the joint equations composed of physical and petrophysical parameters can thus be obtained, and the derivatives of the elastic parameters with respect to petrophysical parameters are obtained by differentiating. The conjugate gradient method is then used to find the optimal solutions of P-wave velocity, shear velocity, density, porosity, water saturation and clay content. We apply this inversion method to prestack seismic data set in Sulige gas field, which is one of the largest gas fields in central Ordos Basin in China. The main gas-producing interval is the H-8 sandstone reservoir at the bottom of Lower Shihezi Formation, Lower Permian, Upper Paleozoic. Drilling results indicate that this region is a composite sandstone gas reservoir and its heterogeneous reservoirs are characterized by low permeability, low pressure and low abundance, which bring great challenge to the exploitation of this field. From the comparison of the inversion result at the well location using our method and two-step method, we can see that there are slight but noticeable differences between these two methods, especially for the elastic parameters. Our joint inversion is significantly closer to the "true" elastic and reservoir model than the two-step inversion. Later we use 2-D field prestack seismic data to assess the performance of the proposed method, and selected a 160 ms time interval of interest for the reservoir description. Three wells that intercept the section are previously characterized in two-way travel time, and used to calibrate the rock physics model and generate low-frequency starting model. The major seismic events in the inversion result show more continuity along the reference structural horizon direction, and the estimated property have a good match with the corresponding well-log curves, and commensurate with the frequency content of the seismic data, as expected. We have developed a method for inverting prestack seismic data under well-log constraints derived from Gassmann fluid substitution relations and prior petrophysical models. In our model, we use Gassmann fluid substitution. relations calibrated to the well-log data for more robust modeling, which is consistent with common petrophysical knowledge. In this specific setting, the extended application to invert prestack seismic data can estimate water saturation, porosity, clay volume, P-wave velocity, shear wave velocity and density jointly. Compared with the stochastic optimization method, our method utilize gradient information of objective function, the optimal solution can be obtained more quickly in favor of large-scale production mission. Compared with Bosch's poststack deterministic optimization, prestack seismic amplitude have more information varying with the angle of incidence. And compared with Wyllie equation, Gassmann equation is better for establishing the relationship between the reservoir parameters and prestack seismic observation data, which helps a lot to understand the impact of reservoir parameters on the elastic properties.