| Reservoir model is an important research tool for oilfield development,and it has an important guiding role for remaining oil prediction and drilling fracturing design,in order to obtain a more accurate reservoir model,it is necessary to combine various aspects of information.Time-lapse seismic reservoir monitoring technology can realize the transition from static description to dynamic characterization of reservoirs,and is an effective method to improve oilfield recovery.Therefore comprehensive time-lapse seismic data can better predict remaining oil and characterize reservoir dynamics.This paper explores the application of time-lapse seismic in reservoir models through rock physics modeling,reservoir model forward modeling,and time-lapse seismic inversion methods.Firstly,according to the basic characteristics of the Angola A block reservoir,a corresponding soft sandstone model is established,but it is also considered that the loose sandstones with irregular arrangement characteristics will relatively slide under the action of stress,which is inconsistent with the basic assumption of Hertz-Mindlin contact theory,so tangential stiffness correction factors c and combined parameters R Rare introduced for correction.The logging test results show that the corrected model’s predicted P-wave velocity and S-wave velocity have a better fit with the actual curve.Then,using the rock physical model to synthesize the three-dimensional reservoir model into a difference impedance data volume,and then the post-stack difference seismic data is inverted to obtain the difference impedance of the entire work area.Finally,the results obtained from reservoir simulation and difference inversion are projected onto the same grid and time-deep domain conversion is performed.Through comparative analysis,it is possible to further clarify the changes of the reservoir before and after development,thereby predicting the distribution of remaining oil.Finally,this paper also studies the time-lapse seismic reservoir parameters inversion,which is mainly divided into two parts.The first is the pre-stack elastic parameters inversion.The accurate Zoeppritz equation is used to establish the forward model,which avoids the error caused by the approximation.In the Bayesian framework,block constraints are introduced to better characterize the boundary,thereby improving the accuracy and stability of the inversion,and then inversion based on rock physics model to predict changes in porosity,pressure and water saturation,so as to characterize reservoir dynamics. |