Le gisement de Rhourde el Baguel (Algérie) : modélisation de la fracturation naturelle
The Rhourde El Baguel Field in Algeria: Modeling of Natural Fracturing
1
Institut Français du Pétrole
2
Université Toulouse III
3
Sonatrach
Le Gisement de Rhourde el Baguel produit de l'huile à partir d'un réservoir gréseux du Cambro-Ordovicien aux qualités matricielles médiocres. La fracturation naturelle joue un rôle important sur la productivité. L'intensité de la fracturation paraît liée davantage aux paramètres lithologiques qu'à des variations d'épaisseur des bancs. Les directions des fractures relevées sur les carottes orientées d'un sondage ne sont pas en accord avec la direction des grandes failles proposée jusqu'ici. Les fractures et les blocs délimités par celles-ci sont présentés sous trois formes : un bloc diagramme simplifié qui montre la complexité du réseau et la faible représentativité d'un sondage à l'échelle du gisement; des blocs matriciels en trois dimensions à partir de l'utilisation d'un logiciel préexistant; une hauteur moyenne de bloc matriciel calculée à partir de l'application d'un modèle probabiliste.
Abstract
The Rhourde el Baguel field in Algeria is made up of a thick series of Cambro-Ordovician sandstone. This highly fractured formation can be subdivided into four terms: R3, R2, Ra and Ri (Fig. 2). Its structure has been interpreted as a horst bounded by two major faults trending NNE-SSW. The western fault has been traversed by well Rb 18, from which the 45° tilted sandstone core samples have been the subject of a special analysis. Different methods were used to evaluate the fracturing. The first consisted in linking the thickness of the beds with the intensity of fracturing (Aguilera method). This method was applied to 700 meters of core samples coming from four wells. Generally, it was found that the thicker a bed was, the more it was fractured, which runs counter to the commonly accepted theory. This finding might be linked to the lithology (the thinner the beds are, the larger the argillaceous matrix content is, and the thicker the argillaceous interbeddings are). Another methodology, based on an analysis of fracturing and the use of probabilistic models linked to the Poisson process, was implemented for well Rb 18, for which dipmeter measurements were available. Fracture directions (Fig. 9) are mainly E-W, which was not expected from the structural pattern accepted previously (Fig. 3). From measurements made in this well, three types of fracture network plots were made :(a) A three-dimensional plot in the form of a block diagram (cf. 2D illustration in Fig. 10). (b) A three-dimensional plot using a software (I-DEAS) (Fig. 11) enabling the blocks to be isolated easily. (c) A plot making use of a Poisson probabilistic model. The last two plots were used to evaluate a mean block thickness of about 45 cm. To conclude, this study illustrates a methodology for analyzing fracturing by using core samples, with these core samples having first been oriented. Since the tilting of the beds could be dated, a relative chronology could be determined for the activity and reworking of the fractures. Different approaches were also used to define the geometry of the matrix blocks having a mean thickness of 45 cm and to reveal the role of lithological parameters in the distribution and intensity of fracturing.
© IFP, 1992