Understanding the spatial distribution of reservoir properties such as lithology and porosity is essential for development drilling, reserves estimation and fluid flow simulation. However, the data typically come from various sources at various scales and have varying degrees of reliability. Data such as wells logs and cores on their own are generally not adequate to produce an accurate model of a reservoir. Geostatistics provides a means for geologists and engineers to analyze this data, and to transfer the resulting analyses and interpretations for the purpose of reservoir modelling and forecasting.

    The objective of this paper is to assess the added value that is gained by integrating different types of data (such as depositional facies and seismic impedance) with 3-D geostatistical porosity models. To achieve this goal, four porosity models of the Hanifa Reservoir at the Berri field (Saudi Arabia) were built using different geostatistical modelling algorithms. The first porosity model was based solely on porosity logs from wells. The other three porosity models were generated using different combinations of porosity logs, depositional facies and seismic impedance data. These models were evaluated qualitatively and quantitatively.

    The results of this study show that facies-based porosity models result in the better definition of porosity both vertically and laterally compared to the other models. The seismic-controlled model was the most precise — seismic data has a greater sample density than well data. Porosity from the wells-only model has the lowest accuracy compared to the other models, which shows the importance of introducing other types of data in porosity modelling.

    It is concluded that the utilization of different data sources has a pronounced positive impact when modelling areas with low sampling density. Integrating seismic impedance and facies data in porosity modeling improves the overall model accuracy and generates more reliable images about reservoir heterogeneity.