This paper discusses approaches to obtaining information on the hydraulic geometry of fracture networks from field data. The approaches include flow rate logging, analysis of packer test transmissivity distributions, and transient analyses of well test data to infer network geometry. Although most well test analyses assume cylindrical flow (spatial dimension of two), the actual dimensions derived from well test data may vary from one to three, including fractional dimensions. Fractional dimensions may be caused by partial connectivity of networks.
Cet article decrit divers moyens d' obtenir des informations sur la geometrie de reseaux de fractures hydrauliquement significatives à partir de donnees de terrain. Les approches considerees incluent le releve des debits à l'Interieur de forages, l'analyse de la distribution des valeurs de transmissivite obtenuesd'essais effectues avec obturateur et l'analyse de donnees d'essais de pompage en regime transitoire dans des puits dans le but de determiner la geometric du reseau de fractures. Bien que la plupart des methodes d' interpretation d' essai effectue dans des puits suppose un ecoulernent cylindrique (espace à deux dimensions), les dimensions derivees de ces donnees peuvent varier d'un à trois, incluant des valeurs fractionnaires de dimension. Les valeurs fractionnaires de dimension peuvent resulter de la connection partielle entre les reseaux de fractures.
Diese Arbeit befaβt sich mit der Bestimmung der hydraulischen Geometrie von Kluftnetzen. Die angefuehrten Verfahren befassen sich mit der Ermittlung und Auswertung von felshydrauilischen Kennwerten aus Einpressversuchen u. dgl. die ueber die verteilung von Duerchlassigkeit, Strömung und Speichervermögen im Gebirge, bzw. Geometrie oles Kluftnetzes Auskumftgeben. Zur Auswertung der meisten Versueche wird eine radiale (zwei-dimensionale) Strömung vorausgesetzt, obwohl die tatsachliche Strömung in mehrefen (1 bis 3) Dimensionen möglich ist. Teildimensionen sind ebenfalls möglich und sind von der teilweisen verbindung die Kluftnetze abhangig.
The study of fracture networks has developed over the past thirty years on the premise that structural geologic information on fracture geometries could be used to develop realistic models of flow. This work has inspired the development of discrete fracture models of fluid flow in rock masses (Long, et al., 1982; Dershowitz, 1984; Herbert and Splawski, 1990). While advances in fracture network modeling have resulted in new capabilities for generating realistic fractures, the simulation of flow in those networks has been hampered by the limitations of computing resources for simulating the large numbers of fractures present in rock masses. Simply stated, it is not generally possible to create a discrete fracture model that includes all of the fractures in the rock mass. The key to efficient network modeling is to limit numerical simulations to the conductive fractures and neglect the remaining fractures or treat them as a matrix porosity in a dual porosity approach. An important task in preparing field data for a discrete fracture model involves identifying the most significant portions of the network for concentrating of simulation efforts. It is widely recognized that relatively small portions of many natural fracture networks control a major portion of the groundwater flow. The characterization of the hydraulic geometry of the fracture network, that is the geometry of the conducting portion of the fracture network, requires a reliance on hydraulic measurements in addition to geophysical and structural geologic data. It is the main purpose of this paper to discuss methods for characterizing the hydraulic geometry of fracture flow systems. The methods described in this paper cover three approaches for defining the hydraulic geometry of fracture networks. The first is the determination of conductive fracture frequency in boreholes by logging and packer test sampling.
