Stochastic Simulation of Lithology Distribution Constrained by Stratigraphic Forward Modeling and Seismic Inversion: A Case Study of A Oilfield in Ecuador

Abstract

Prediction of lithology distribution is critical for development of oilfield and is also one of the basic tasks of geological modeling. Stochastic simulation has been proved to be an effective method for simulation the distribution of lithology. While stochastic simulation technique is theoretically a mathematical method, the result of simulation will be strongly affected by many factors, such as the quantity, quality, and distribution of well data, and so on. In the area or oilfield with limited wells and insufficient data, the uncertainty of simulation result will be very high, and the reliability of the result will be very low. This result with high uncertainty and low reliability will lead to very poor prediction of sand and shale. Further, the reliability of the result of petrophysical modeling will be low. This article introduces a new method by using constraint condition with high quality to improve the simulation result of lithology distribution and decrease the uncertainty of simulation result. Stratigraphic forward modeling is an effective approach to simulate the process of deposition, and this approach can characterize the geometry and distribution of complex sedimentation. But the resolution of the simulation result of this approach is lower than the requirement of development stage of oilfield. Seismic inversion has been proved to be an effective way with high resolution to describe the distribution of lithology by using high-quality seismic data, while seismic inversion is difficult to reflect the geological process and pattern. The combination of these two methods can provide good constraint condition for stochastic simulation. By using the results of stratigraphic forward modeling and seismic inversion, we can build the trend models for different lithology. The trend models can be used in stochastic simulation, and the trend models can improve the result of stochastic simulation. This paper took A oilfield in Oriente Basin in Ecuador as an example to show the process of this method. We built the stratigraphic forward modeling model based on the geological understanding and completed the research of seismic inversion of M1 layer of Napo Formation. The result of serigraphic forward modeling and seismic inversion was resampled into the geological model. Based on these two results, the trend models of sand and shale of M1 layer have been completed. The trend models have been used to constrain the stochastic simulation of lithology with sequential indicator simulation algorithm. This method can effectively decrease the uncertainty of simulation result caused by insufficient well data and improve the reliability of lithological model which further improves the result of petrophysical modeling.