Seismic Constrained Reservoir Simulation and Application in a Heavy Oilfield, China

Abstract With the development of reservoir numerical simulation, especially thermal numerical simulation, people start to use thermal numerical simulation model to study the development and distribution of steam chamber in thermal recovery process. But as all known, there still are many uncertainties in the reservoir parameters and well parameters in thermal simulation model, such as the excessive variation of the fluid properties, the uncertainty of static model owing to reservoir heterogeneity and others. These uncertainties usually result in multiple solutions of simulation history matching or ambiguous simulation results. To reduce multiplicity of history matching in reservoir simulation and better predict the development of steam chamber in the thermal recovery process, seismic constrained reservoir numerical simulation is proposed to incorporate both reservoir simulation and 3D seismic image result in the loop of reservoir simulation and history match. The procedure is discussed through the application to a heavy oil reservoir in which two times of 3D seismic surveys were acquired at the later stage of production, and reservoir numerical simulation was performed in the SAGD (Steam assisted gravity drainage) pilot block of the reservoir. The proposed procedure includes following main steps: 1) The dynamic data is collected and analyzed for the preparation of reservoir numerical simulation, and an overall distribution of steam chamber is predicted approximately through reservoir engineering analysis; 2) A reservoir simulation model is initially built using available geological, petrophysical and engineering data, and primary history matching is then performed; 3)The reservoir simulation results from previous step and the time-lapse seismic imaging are interpreted comprehensively to find out the consistencies and differences about the steam chamber between them, which are related to history matching errors. 4) Based on these consistencies and differences, the possible factors (geological properties, petrophysical parameters, and rock thermal properties) which caused the history match errors are analyzed, and the corresponding parameters of the simulation model are modified accordingly, and then back to step 2 and 3 for the iteration until a relatively matched result proved by subsequent production data was received. Application results proved that with the constraint of seismic image data, the reservoir simulation errors can be accurately located and the primary cause for it be determined, and then through reservoir simulation iterations the relatively accurate prediction of steam chamber as well as remaining oil is reached.