ABSTRACT:

This paper describes recent progress in the development of conceptual models for fracture geometry, and presents field and numerical modelling results which illustrate the importance of an accurate representation of the spatial structure of fracturing. Topics covered include fracture shape, fracture location, fracture intensity, fracture size, planarity, and roughness. Statistical tests are presented to compare field measurements of fracture properties against the parameters of the alternative conceptual models. The importance of defining appropriate conceptual models is illustrated using applications in analysis of flow and deformation of fractured rocks.

ABSTRACKT:

Der Vortrag beschreibt den derzeitigen Stand in der Entwicklung von konzeptionellen Modellen zur Beschreibung der Geometrie von K1uftstrukturen. Es werden Feldbeispiele und die Ergebnisse numerischer Modellierungen vorgestellt, welche die Wichtigkeit einer genauen Darstellung der raumlichen Struktur von Kluftsystemen verdeutlichen. Dieses beinhalt die Beschreibung von Kluftform, Kluftlokation, Kluftintensitat, Kluftgröβe, und Rauhigkeit. Weiterhin werden statistische Tests vorgestellt, die einen Vergleich von Feldmessungen der Kluftporositat mit den Ergebnissen der verschiedenen konzeptionellen Modelle erlauben. Die Wichtigkeit einer korrekten Wahl eines konzeptionellen Modelles wird anhand von Anwendungen bei der Analyse von Fliessvorgangen in Klueften und der Untersuchung von Deformationen gekluefteter Formationen demonstriert.

RÉSUMÉ:

Ce document decrit recentes progres dans le developpernent des modeles conceptuels pour la geometrie de la cassure, et presente un domaine et une moulage numerique qui explique importance d'une representation de la structure spatiale de la fracturation. Les sujets traites comprendent forme, emplacement, intensite, dimension, planerite et rugosite de la cassure. Ils ont ete presente test statistiques pour etablir une comparaison des donnees de campagne correspondants à les cassures avec les parametres des modeles teoriques alternatifs. L'importance de definir modeles teoriques est illustre avec I'applications de l'analyse du flux et deformation de la roche cassure.

1. INTRODUCTION

It is surprising that even today, many discrete fracture models utilize simplified assumptions for fracture geometry, even though the predominance of evidence indicates that more sophisticated fracture patterns are required to provide a realistic representation of reality. Current models assume that fractures are infinite (Cundall and Hart, 1992), circular (Cacas, 1991), or rectangular (Herbert, 1992). Fracture locations are assumed to be located according to a Poisson process (op. cite.) Statistical evaluations of field measurements clearly demonstrate that these assumptions are not adequate. This paper describes the conceptual models which are currently available for fracture geometry. These models should be used whenever discrete fracture approaches are used.

2. FRACTURE LOCATION

For rock mechanics problems of flow, deformation, and stability, the connectivity of fractures is frequently the key parameter (Figure 1). The connectivity of fracture patterns is controlled by fracture location, together with fracture size and intensity. As a result, it is essential to develop the correct conceptual model for fracture location. A series of seven models have been implemented and evaluated in three dimensions within the FracMan discrete feature modelling system (Dershowitz et al., 1992).

2.1 Poisson Location Model ("Baecher Model"):

In the Baecher model, fracture locations are assumed to be defined by centers uniformly located in space. This assumption is tested by overlaying a grid on the fracture pattern, and comparing the distribution of the number of fracture centers per grid cell against the Poisson distribution using the Chi-Squared test.

2.2 Multiple Poisson Location Models ("BART Model"')

In multiple Poisson models, the fracture location is defined by a Poisson process of location seeds from previously defined fractures, rather that from seeds located is space (Figure 2). This results in a strong correlation between fracture locations, and a higher fracture connectivity. These models can be evaluated by fitting correlation functions between fracture centers, and also by calculating probability that fracture terminations are found at fracture intersections rather than in intact rock.

This content is only available via PDF.
You can access this article if you purchase or spend a download.