Development of an objective function to measure the dissimilarity between data observed in wells and results of Stratigraphic Forward Models-SFM

Resumo

In the area of oil production investment decisions and production planning depend on oil production
predictions, which are made using reservoir simulation techniques. Simulations reproduce the flows of
hydrocarbons in the reservoirs, described through a geological model. Geological models integrate data (well,
seismic) and geological interpretations. Stratigraphic Forward Models (SFM) consists in modeling geological
processes that considers the principles of conservation of mass, energy and reproduces processes in the study
domain (sedimentary basin). SFM provides the facies distribution in the sedimentary basin, resulting from the
geological processes. Once the SF modeling has been carried out, it is necessary to evaluate the results obtained
and apply a calibration since the results do not honor information from the well data. In order to solve the inversion
problem, several authors have developed different methodologies. These approaches provide numerous
contributions of extreme relevance, but none of them presents a possible solution for operational application due
to an incomplete or inadequate configuration of the objective function (generally not sufficiently geologically
representative). Therefore, this work describes a possibility more adequate objective function that clearly and
quantitatively translates the degree of similarity between the numerical results and the data is one of the main keys
of a successful inversion. The proposed objective function was developed based on previous work, which used an
approach originated from automated well correlations. Some modifications of this method were proposed to make
the objective function more adapted to SFM calibration. Among these modifications are the fact that the objective
function can now compare SFM results whose resolution differ greatly from the resolution of the calibration data.