ABSTRACT: An experimental study of a low strength fluvially deposited triassic sandstone has been performed. The investigation correlates the anisotropy in elastic (static and dynamic) moduli with that in strength parameters. Also, permeability anisotropy is measured. Microscopic studies have been performed to relate the observed mechanical anisotropy to the rock's microlayering. The results are compared with theoretical models. Some consequences of rock mechanical anisotropy for borehole or cavity stability are briefly commented.
RESUME: Une etude experimentale d'un faible grès d'age Trias a ete effectuee. L'etude fait une correlation entre l' anisotropie des paramètres elastiques (statique et dynamique) et les parametres de resistance. L'anisotropie de permeabilite est aussi mesuree. Des etudes microscopiques ont ete effectuees afin de faire une relation entre l'anisotropie mechanique et la nature des microcouches de la roche. Les resultats sont compares avec les modeles theoriques. Quelques effects de cette anisotropie mechanique sur la stabilite d'un puits ou d'une cavite sont brevement discutes.
ZUSAMMENFASSUNG: Anhand von Experimenten mit fluvialen, triassischem Sandstein von geringe Festigkeit wird versucht, eine Verbindung zwischen der Anisotropie der statischen und dynamischen Moduli und verschiedenen Festigkeitsparametern unabhangig vom Umgebungsdruck herzustellen. Mikroskopische Undersuchungen sollen die Beziehung zwischen der beobachteten mechanischen Anisotropie und der Feinschichtung des Gesteins beleuchten. Die Resultate werden mit theoretischen Modellen vergleichen und einige Konsequenzen der mechanischen Anisotropie fuer Bohrloch- und Kavernenstabilitat werden kurz erwahnt.
INTRODUCTION Rock mechanical problems associated with weak sedimentary rocks are often encountered by the petroleum industry. Instabilities may occur during drilling and production and may be located to the well itself (breakouts, sand production etc.) or the entire reservoir (reservoir compaction, surface subsidence). In addition, also evaluation of stability of excavations and tunnels in sedimentary strata onshore or offshore call for an improved understanding of the mechanical behaviour of weak rocks. There is also a need for procedures by which mechanical properties can be evaluated from borehole measurements. Most petroleum reservoirs produce from sandstone formations. When evaluating well stability, current models (Coates and Denoo, 1981; Geertsma, 1985) are based-on empirical correlations established from a limited set of experiments with rocks from a limited geographical area. Most of the materials used in these correlations are significantly stronger than the formations causing for instance sand problems. Also, isotropy is generally assumed. At least for sandstones, anisotropy is thought to be of minor importance. Rock bedding may, however, have consequences for the stability of boreholes. Breakouts are formed in collapse situations, elongating the hole in the direction of the minimum in-situ horizontal stress. Clearly, strength anisotropy may influence on the breakout direction. In particular, if the horizontal stress components are equal, the direction of the breakout will be given by the bedding planes. When dealing with deviated holes (Aadnoy and Chenevert, 1987), the borehole inclination will have to be chosen with care in order to optimize the stability. When drilling through a horizontally bedded formation, the most sensitive range of inclinatian with respect to borehole collapse is 10–40°, when the rock will fail along the bedding plane. This paper presents results from a series of rock mechanical experiments performed on a weak sandstone in order to characterize its anisotropic mechanical behaviour with respect to strength parameters and static as well as dynamic elastic moduli.
MATERIAL CHARACTERIZATION The sandstone used in this study was sampled from an outcrop in Great Britain. We have given it the name "Red Wildmoor sandstone" (RW). The rock is of Triassic origin, and was fluvially deposited. The mineralogy is quite homogeneous throughout the sampled material. The homogeneity is confirmed by He porosity measurements, which give a more or less constant porosity of 26±1%. The main constituents are quartz with some feldspar and fragments of crystalline quartzites and gneisses. Small amounts of mica and clay minerals like illite and kaolinite can be found. The colour is red due to ferrous oxide, possibly hematite mixed with clay minerals, surrounding the grains as a thin film. Neither carbonate nor quartz cement have been observed. The texture may change significantly from block to block. Figures 1 and 2 illustrate two different structures: In Figure 1, a thin section photograph of material type A, is shown. Areas with a clearly visible lamination can be found. The grains are subrounded to rounded, intermixed with some few angular grains. The sphericity is higher than average, while the orientation of elongated grains seems to be random. The grain to grain contacts are primarily tangential. Grain sorting is poor to moderate. Both fine (grain size 80–100 11m) and coarser grained regions (bimodal grain size distribution, the larger grains being 220–230 11m and the finer fraction 60–100 11m in diameter) can be found. These regions are separated by lamina of silt having the same mineralogy as the sand fractions, but with a richer content of clay minerals. Figure 2 shows thin section photographs of material type B.