Online Journal for E&P Geoscientists

Nofli, Sami ^*1
(1) PDO, Muscat, Oman.

The degree of heterogeneity is one of the most important geological uncertainty parameters as it has a significant impact on field development. The significance of small-scale heterogeneity becomes even more apparent in heavy-oil fields undergoing EOR (thermal) development.

In the studied field, the reservoir is composed of high net-to-gross (>90%) fluvial sediments that can be divided into smaller fining-upwards units, but cannot be laterally correlated. The porosity logs show a uniform behaviour throughout most of the vertical section in the oil zone interval. However, the Gamma Ray and the resistivity logs suggest that the system preserves some small-scale heterogeneity. Comparison between the petrophysical logs and the core data indicates that thin (< 30 cm) mudstones are present which are below log resolution and which can form baffles to flow. Other types of heterogeneity in this fluvial system are represented by cemented beds and fractures, mudclasts, and possibly grain size variations.

The modelling challenges of such a high net-to-gross system with associated small-scale heterogeneity are to understand its lateral extent and the 3D distribution and to capture and preserve the vertical distribution during the up-scaling process.

Facies Modelling, in particular baffle modelling, is very essential as the baffles have a major effect on the vertical permeability. An attempt was made to capture the thin mudstone beds, using the shallow resistivity logs. The shallow resistivity approach, along with the GR is used to manually define the thin mudstone facies. Defining the dimensions and the densities of the baffles away from the well-bore is another modelling challenge. With the available information and current understanding of possible baffle dimensions, a geological realisation with the 0.5 m layering scheme is performed. The baffles are modelled and distributed in a 25 m by 25 m 3D grid, using SIS (Sequential Indicator Simulation).

The facies modelling will be improved after incorporating the output of the high-resolution FMI study, which aims to identify, quantify and correlate the baffles with the core data.