Click to view figure in PDF format

Reservoir Characterization and History Matching of a Fractured Carbonate and Highly Permeable Sandstone Reservoir

Tone Nedrelid¹*, Kjersti Håland¹, Kellfrid B. Lyslo¹, Mehran Azizzadeh², Javad Honarmand², Ghorbanali Sobhi², Michael Hovdan¹, Claus von Winterfeld¹**.
¹Statoil ASA, 4035 Stavanger, Norway.
²RIPI, Tehran, Iran.
* [email protected]
**Present address: Petroleum Development Oman LLC, Muscat, Oman

The Field is located near the Persian Gulf at the foothills of the Zagros Mountains. The reservoir consists of interbedded limestones, dolomites and clastic sediments.

The aim of the study was to build a 3D reservoir model describing the matrix properties and the fracture distribution in the reservoir. Further, the aim was to generate a history matched dual porosity reservoir simulation model, from which predictions of improved oil recovery could be made.

A new reservoir zonation built on a sequence stratigraphic framework was developed to improve the modelling of flow units in the field. A 3D corner point grid was built in RMS with near orthogonal grid cell geometry. The grid is laterally identical with the simulation grid, only the vertical resolution differs. The 3D grid was populated with stochastic realizations of a well and trend constrained facies model. The distributed facies objects (dolostone, limestone, sandstone, shale) were then populated with their specific petrophysical properties again conditioned to well data and observed spatial data trends.

Based on static geological and dynamic well data in addition to outcrop analogue data, a conceptual fracture model was established, representing the general understanding of the distribution and formation of fractures. Using this conceptual fracture model, a 3D discrete fracture network (DFN) model was generated using FRACA^TM. The DFN model was up-scaled to Full-Field Equivalent fracture parameters for the ECLIPSE model.

The reservoir simulation model was tuned to match the initial highly tilted oil-water contact created by a dynamic aquifer (pre-production) and tuned to match the overall energy distribution suggested by material balance.

Matching criteria for the production history were set for three main issues:

• Pressure development
• Produced Gas-Oil-Ratios
• Contact movements

These criteria made it easy to identify deviations and problem areas, and the worst offending matters were continuously improved.

The result was a high quality history matched dual porosity reservoir simulation model.

Figure 1