Improved Stochastic Simulation Using Stratigraphic Forward Modeling: A Case Study of the Lithological Distribution of Tide-Dominated Estuary in JE-AW Oil Field, Ecuador

Abstract

Stochastic simulation of lithology is critical in geological modeling, and the result quality relies on the well data analysis. The traditional method may involve too many uncertainties due to the limited wells. This paper introduces a new method to improve the simulation of lithology distribution of tide-dominated estuary using probability volume generated by stratigraphic forward modeling as key input for trend model. Based on the core analysis and investigation on regional geology, the stratigraphic forward modeling approach generates an initial model by using the sequence stratigraphic scheme. And the sensitivity analysis provides an indication of adjusting the influencing parameters which control the sand and shale distribution. The models will be compared with the well data, geological concept, and seismic attribute. The selected model will be resampled into geological grid to generate the trend volume combined with seismic inversion data. Further, the lithology distribution can be simulated by using stochastic method with the trend volume. This approach has been successfully applied in JE-AW oil field to improve the geological model of M1. Typical tidal sedimentary structures, such as mud drapes, and wavy bedding shown in core and overall upward fining shown in logs, reveal the tide-dominated estuary environment during the deposition of M1. Three sub-zones (layers) of M1 are identified and correlated. Based on the sensitivity analysis, the sediment input and subsidence is adjusted for reliable stratigraphic forward modeling. The trend model is generated by inputting the result from stratigraphic forward modeling and seismic inversion. Finally, the lithology distribution is simulated using the trend model. This method improves the lithology stochastic simulation of tide-dominated estuary honoring the well and seismic data. This method reduces the uncertainties of stochastic modeling caused by limited wells and improves the predictability of lithology model.