Back to results
US 12,282,128 B2Granted
System and method for quantitative seismic integration modeling workflow
AssigneeSaudi Arabian Oil Company
InventorsMuiz A. Elmardi Widatalla, Aiman Malike Bakhorji, Mukarram Ahmed
Life of this patent
Estimating the term…
Abstract
Systems and methods for quantitative seismic integrated modelling (QSIM) are disclosed for integrating the one, two and three-dimensional (1D, 2D, 3D) data from different geoscience domains within a framework in order to produce hi-resolution geocellular models that simulate realistic sub-surface reservoir properties. The QSIM systems and methods accurately leverage the seismically derived reservoir rock properties, integrating the geophysical, geological and engineering information through an optimum rock physics models and takes in consideration all the empirically constrained templates to correct, validate and quality check all the input data.
Claims20
A computer implemented method of modelling a subsurface region of the earth for hydrocarbon exploration, development, or production, the method comprising:
- receiving into a processing system pre-stack seismic datasets measured for one or more locations within the subsurface region and well log datasets measured from wells within the subsurface region, wherein the seismic datasets comprise three-dimensional (3D) seismic amplitude data and wherein the well log datasets comprise core data and elastic logs;
- selecting, with the processing system, key wells from among the wells based on the well log datasets;
- generating, based on the seismic datasets, a geophysical model, wherein the geophysical model is 3D and wherein generating the geophysical model includes:
- measuring, based on the seismic amplitude data and the well log datasets for the key wells, an amplitude variation with offset (AVO) response of the seismic amplitude data to determine whether the measured AVO response meets prescribed requirements, wherein the step of measuring AVO to determine whether the measured AVO response meets prescribed requirements includes:
- performing a full waveform inversion process to generate synthetic seismic gathers using wavelets estimated from original input seismic data, convolved with reflectivity calculated from the elastic well logs, and
- comparing, a measured AVO response from the synthetic seismic gathers with a measured AVO response of the processed seismic amplitude data to determine whether the measured AVO response meets prescribed requirements,
- in response to determining that the measured AVO response does not meet the prescribed requirements, iteratively changing at least one seismic amplitude data processing parameter and re-processing the seismic amplitude data and repeating the step for measuring AVO at least until determining that the AVO response meets the prescribed requirements, and
- performing a stochastic inversion process on the re-processed seismic amplitude data;
- generating based on the well log datasets, a geological model, wherein generating the geological model includes:
- defining core facies from the core data for the key wells,
- generating, based on the elastic logs for the key wells and according to the stochastic inversion process of the re-processed seismic amplitude data, an elastic facies model for predicting elastic facies from elastic logs, and
- associating the elastic facies with the core facies;
- generating a conceptual depositional model of the subsurface region;
- and
- integrating the generated geophysical model, geological model and conceptual depositional model into a 3D reservoir model for the subsurface region using a stochastic integration and using probabilistic models to quantify uncertainties of the 3D reservoir model by generating multiple realizations of the reservoir model capturing multiple possible scenarios, wherein the stochastic inversion process on the re-processed seismic amplitude data is performed using a same 3D grid as the 3D reservoir model, wherein the 3D grid is in a depth domain,
- wherein the stochastic inversion process is performed in the depth domain on the re-processed seismic amplitude data, wherein the stochastic inversion process applies Bayesian interference and Markov Chain Monte Carlo sampling algorithms using core data and elastic logs for the key wells and a velocity model linking the depth domain and a time domain of the seismic datasets to generate the geophysical model in the depth domain,
- wherein the step of generating the elastic facies model comprises:
- generating, using elastic log data for the key wells and statistical analysis techniques, a reference elastic facies model defining the elastic facies,
- associating the defined elastic facies with the core facies for the key wells,
- applying the reference elastic facies model to elastic logs for non-key wells among the wells to predict elastic facies for the non-key wells, and
- associating the predicted elastic facies with core facies from the non-key wells;
- wherein the steps of generating the elastic facies model, generating the geological model and generating the geophysical model are performed in parallel;
- and
- updating the elastic facies model, the geological model, the geophysical model, and the integrated 3D reservoir model using based on one or more of:
- new pre-stack seismic datasets measured within the subsurface region, and new well log datasets measured from wells within the subsurface region.
22 more elements
A system for modelling a subsurface region of the earth for hydrocarbon exploration, development, or production, the apparatus comprising:
- a computing device including:
- a non-transitory storage medium,
- an interface for receiving pre-stack seismic datasets measured for one or more locations within the subsurface region and well log datasets measured from wells within the subsurface region, wherein the seismic datasets comprise three-dimensional (3D) seismic amplitude data and wherein the well log datasets comprise core data and elastic logs,
- a processor, wherein the processor is configured by executing code that is in the form of one or more modules stored in the non-transitory storage medium, wherein the one or more modules configure the processor to:
- select key wells from among the wells based on the well log datasets,
- generate, based on the seismic datasets and the well log datasets and according to the stochastic inversion process of the re-processed seismic amplitude data, a geophysical model, wherein the geophysical model is generated by:
- measuring, based on the seismic amplitude data and the well log datasets for the key wells, an amplitude variation with offset (AVO) response of the seismic amplitude data,
- determining whether the measured AVO response meets prescribed requirements, wherein determining whether the measured AVO response meets prescribed requirements includes,
- performing a full waveform inversion process to generate synthetic seismic gathers using wavelets estimated from original input seismic data, convolved with reflectivity calculated from the elastic well logs, and
- comparing, a measured AVO response from the synthetic seismic gathers with a measured AVO response of the processed seismic amplitude data to determine whether the measured AVO response meets prescribed requirements,
- in response to determining that the measured AVO response does not meet the prescribed requirements, iteratively changing at least one seismic amplitude data processing parameter and re-processing the seismic amplitude data and repeating the step for measuring AVO at least until the AVO response meets the prescribed requirements, and
- performing a stochastic inversion process on the re-processed seismic amplitude data,
- generate based on the well log datasets, a geological model, wherein generating the geological model includes:
- defining core facies from the core data for the key wells,
- generating, based on the elastic logs for the key wells, an elastic facies model for predicting elastic facies from elastic logs, and
- associating the elastic facies with the core facies,
- generate a conceptual depositional model of the subsurface region based on the well log datasets, and
- integrate the generated geophysical model, geological model and conceptual depositional model into a 3D reservoir model for the subsurface region using a stochastic integration and using probabilistic models to quantify uncertainties of the 3D reservoir model by generating multiple realizations of the reservoir model capturing multiple possible scenarios, wherein the stochastic inversion process on the re-processed seismic amplitude data is performed using a same 3D grid as the 3D reservoir model, wherein the 3D grid is in a depth domain and the stochastic inversion process is performed in the depth domain on the re-processed seismic amplitude data, wherein the stochastic inversion process applies Bayesian interference and Markov Chain Monte Carlo sampling algorithms using core data and elastic logs for the key wells and a velocity model linking the depth domain and a time domain of the seismic datasets to generate the geophysical model in the depth domain,
- wherein the processor is configured to generate the elastic facies model by:
- generating, using elastic log data for the key wells and statistical analysis techniques, a reference elastic facies model defining the elastic facies,
- associating the defined elastic facies with the core facies for the key wells,
- applying the reference elastic facies model to elastic logs for non-key wells among the wells to predict elastic facies for the non-key wells, and
- associating the predicted elastic facies with core facies from the non-key wells,
- wherein the processor is configured to generate the elastic facies model, generate the geological model and generate the geophysical model in parallel, and
- wherein the processor is configured to update the elastic facies model, the geological model, the geophysical model, and the integrated 3D reservoir model using based on one or more of:
- new pre-stack seismic datasets measured within the subsurface region, and new well log datasets measured from wells within the subsurface region.
23 more elements
Dependencies are inferred from the claim text.