Logo GEPlan

AUTHORS: R. Di Cuia* , A. Riva & M. Marian (G.E.Plan Consulting, Italy), D.Casabianca (Marathon International, UK), E. Forte (Univ.of Trieste, Italy)

YEAR: 2010

Presented at Geo2010 conference in Manama (Bahrain) on March 2010

Fractured Carbonate reservoirs are liable to display large variability in their characteristics which can affect their performance and economic viability. Primary facies distribution and properties, sequence stratigraphic framework, diagenesis and fracturing are amongst the main factors that exert important controls on reservoir properties. In these reservoirs the interaction between matrix characteristics (facies, layering, poro-permeability, wettability) and fractures characteristics (timing, style, sizes, distribution, orientation) invariably control fluid flow by enhancing or imparting primary reservoir properties. As a result the understanding of fracture distribution and in particular the relationship between fractures and sedimentary facies is fundamental for an adequate description of fractured carbonate reservoirs. Understanding these relationships is key to correctly model matrix storativity and fracture connectivity and identify potentially highly productive intervals.
To help to constrain the modelling of these kind of reservoirs the geoscientist must make use of outcrop analogues which supply detailed 3D information about all the key parameter impacting the reservoir performances.
We have selected a beautifully exposed 3D outcrop to test an innovative integrated methodology using geological (structural, sedimentological, diagenetic) studies with a geophysical data acquisition (a 3D survey and some 2D lines using the GPR– ground penetrating radar ). The result is a comprehensive model that highlights at small scale the resolution and pitfalls of the different methodologies and supply important information and constrain on how to correctly integrate data coming from different sources for building a reservoir model. The lack of integration between data increases the uncertainties on the model produced