A detailed reservoir characterization is an essential requirement for a reliable prediction of flow behavior and effective recovery processes in hydrocarbon exploitation and production. An important aspect in reservoir characterization is the geometric description of sedimentary facies and permeability architecture. These structures commonly presenting local heterogeneities, that is local changes in directions as well as anisotropy, requiring non-stationary processes for their description. Traditional geostatistical pixel-based algorithms do not allow for varying directions of anisotropy. consequently fail to reproduce local structural changes of sedimentary facies and reservoir properties.
This work presents a new geostatistical technology that allows an integration of information from different sources, description of local and global structural directions and prediction/characterization of reservoir architectures and flow properties, leading to a more accurate prediction of spatial continuities, drainage areas, flow capacity and finally optimum exploitation scheme.
The methodology and solution technique as well as their applications are described and illustrated through case studies in the North of Monagas fields, Eastern Venezuela.
Architecture and internal structures of reservoirs are key factors with strong influence in fluid flow and hydrocarbon exploitation and production. Their thorough knowledge are essential for accurate prediction of fluid migration paths and reservoir behavior. In addition, we can design and implement more realistic recovery strategies, that account for the internal reservoir compartments, and other potential characteristics such as lithology, porosity and permeability distribution within each compartment.
This paper outlines briefly first the concept behind of non- stationary cell-based geostatistical techniques, originally pro posed by Stanford Center for Reservoir Forecasting Consortium, and then present a case study with real data from North of Monagas fields in Eastern Venezuela (Fig. 1). This area is one of the most prospective and prolific zone with potential hydrocarbon accumulations operated by CORPOVEN, S.A., one of the PDVSA's affiliates. It conforms 3500 feet of productive section with sediments deposited from Cretaceous age to lower Miocene. These reservoir are actually under extensive study to compliment the development plan aiming to implement high pressure gas injection program yet to optimize recovery efficiency.
In a detailed characterization campaign that is under way in a joint project between INTEVEP and CORPOVEN, we have proved several emerging technologies. Geostatistical techniques with local direction of anisotropy is one of them, that were applied in a pilot area of North of Monagas. In this paper we present partial results of this study.
Conventional cell-based geostatistical methods of estimation and simulation do not allow for varying directions of anisotropy and fail to reproduce the local architectural variation present in a given reservoir. This call for a non-stationary techniques that are achieved by modifying existing algorithms for estimation or simulation of continuos and categorical variables (Fig. 3).
The steps involved in this modeling process are three fold:
Estimation or simulation and mapping of directional angles (u), by traditional methods. This define the local direction of continuity at each location u, conditioned to the control data values (ua) where, = 1, . . . , n.
