Modelamento estocastico integrando dadso de poços horizontal e verticais

AUTOR(ES)
DATA DE PUBLICAÇÃO

1993

RESUMO

This thesis presents a stochastic model for the spacial distribution of effective porosity of heterogeneous fluvial reservoirs. The 500 m-wide and 800 m-long study area contains 19 vertical wells and a horizontal well with a 600 m-long "3D" trajectory. A 18 m-thick portion of the reservoir was investigated; this section is defined by the vertical oscillations of the horizontal well. The studied portion of reservoir was divided into blocks with dimensions of20 x 20 xl m or 18,000 cells. This procedure allowed the vertical resolution for simulation required by the large variability in effective porosity presented by the. reserve. Effective porosity was estimated by log-rock correlations, and oil-saturated horizons were constrained by a porosity cut-off of24%. Two types of simulation were developed: (1) simulation based on a variographic model calculated with information from vertical wells only; this simulation tends to show great porosity values with more continuous spatial distribution; (2) simulation based on a variographic model calculated with information from both the vertical and horizontal wells; in this case the estimated poro sities show a larger variability and a more complex distribution. All of these simulations were developed with the same randomic seed. The applied stochastic models were used in the simulation of a continuous variable (effective porosity), with a basic assumption of ordinary krigging. The three types of -simulation were compared to three horizontal views (slices) of the reservoir, respectively located at 1 m, 10 m, and 18 m bellow the top of the reservoir. In the third simulation listed above, a total of six vertical views (cross sections) of the reservoir were obtained. Horizontal wells provide a more detailed description of the spatial variability of reservoir properties, supporting the development of more realistic variographic models; these, in turn, tend to improve reservoir stochastic simulations, by reducing the inferences involved in theoretical models

ASSUNTO(S)

areservatorios analise estocastica pocos de petroleo

ACESSO AO ARTIGO

Documentos Relacionados