Search

A field-specific geomechanical model serves as a platform for greatly reducing costs and increasing production over the life of a field. The information contained in a geomechanical model makes it possible to reduce drilling costs and production losses through fieldwide well planning that can optimize production and minimize risk. A significant value of the geomechanical model is its application to the efficient exploitation of fractured reservoirs. The essential contribution of wellbore image technologies to this exploration and production challenge is illustrated through a case study of a compartmentalized fractured gas reservoir located in Hokkaido, Japan. A growing body of evidence reveals that, in many fractured reservoirs, the most productive fractures are those that are optimally aligned in the current stress field to fail in shear. Thus, it is necessary to obtain knowledge of both the stress magnitudes and orientations and the distribution of natural fractures to determine the optimal orientations for wells to maximize their productivity. The best well intersects the maximum number of stress sensitive fractures. Applying geomechanics and the reservoir fracture distributions to model shear-enhanced permeability as the mechanism for reservoir production appears to be a promising improvement to existing reservoir flow models. Using quantitative risk assessment and realistic uncertainties in the critical parameters, it is possible to estimate the uncertainty in predictions of optimal well trajectories and of stimulation pressures to enhance natural fractures. The results indicate that the critical parameters are not always those with the most uncertainty, and that the most effective way to reduce prediction uncertainties is to calibrate against the productivity of a preexisting well.

Gassmann’s (1951) equations relate the elastic bulk (Ksat) and shear (Gsat) moduli of a fully saturated rock to the elastic bulk and shear moduli of the dry‐rock frame (Kdryand Gdry, respectively), porosity ϕ, and the bulk moduli of the mineral phase [Formula: see text] and pore fluid [Formula: see text].

Recent seismic experiments have revealed that a shallow, low velocity zone is a common feature within the uppermost few hundred meters of very young (less than 100 kyr) oceanic crust at ‘fast spreading’ ridges. The shape of this zone suggests that the low velocities are related to the constructional morphology of the extrusive (layer 2A) portion of the crust. With this in mind, we use simple analytical techniques to study the effects of characteristic morphologic elements of the extrusive basalts on their elastic properties. We find that high compliance of the contacts between pillows and pillow fragments can account for all of the observed velocity anomaly. Voids between pillows are, by comparison to their effects on porosity, considerably less important. Radial (cooling) cracks and voids within the pillows themselves will affect the velocities in direct proportion to their effect on the individual pillows. The relatively rapid increase in velocity with age in these very young extrusives can be explained by welding of pillow contacts and stiffening of the individual pillows by infilling of radial fractures, both of which may take place without a large reduction in porosity. If voids between pillows are filled initially by sediments, the cementation of these materials can also cause large increases in velocity without a large addition of mass. Because of the importance of such subtle variations in morphology, there can be no one-to-one relationship between seismic velocities and porosities within the extrusive layer of the oceanic crust. Therefore independent measures of crustal porosity must be found (for example, using sea floor gravity surveys).

Ultrasonic compressional and shear‐wave velocities of isotropic sands are shown to be dependent on their mineralogy, their porosity, their fluid content, and their state of consolidation, under fixed temperature and pressure conditions. This leads to a distinction between two broad classes of sands: those that are well consolidated, and those that are loosely consolidated. Changes in elastic velocities reflect changes in the ratio of bulk and shear moduli to density in response to lithologic variations. We decouple the two effects by examining changes in elastic moduli with respect to changes in lithology, and we observe three main points: (1) For consolidated sandstones, the effects of mineralogy and porosity can be approximated both empirically and theoretically by a modified isostrain theory: the dry bulk and shear moduli of the rock aggregate follow a “mixing law,” being linear combinations of the respective moduli of the individual constituents. The dry elastic moduli of families of clean sands and shaley sands are linear functions of porosity, with decreasing y‐axis intercepts as their clay‐to‐sand ratio increases. (2) Loosely consolidated sands and sandy shales appear to follow a behavior closer to that of the isostress theory for suspensions: the reciprocals of the bulk and shear moduli of the rock aggregate are linear combinations of the reciprocal moduli of their individual constituents. In general, the elastic moduli of poorly lithified sands are less sensitive to changes in mineralogy and porosity than those of consolidated sandstones. (3) For high permeability sands like the loosely consolidated sands of Troll, the Biot‐Gassman theory is a good approximation to the effects of fluids on seismic velocities. With our understanding of elastic moduli, we then show that dry ratios [Formula: see text] increase with porosity and clay content.

Multichannel full waveform acoustic logs were recorded in Deep Sea Drilling Project (DSDP) hole 418A during Ocean Drilling Program (ODP) leg 102, DSDP hole 395A during ODP leg 109, and DSDP hole 504B during ODP leg 111, to provide nearly continuous measurements of elastic wave velocities as a function of depth within oceanic layer 2 for different spreading rates and crustal ages. The velocities depend primarily on the morphology of the basalt. Massive units have Vp above 5 km/s, Vs above 2.8 km/s, and Vp/Vs below 1.9. Their velocities increase with depth as expected from crack closure due to confining stress and are lower where fractures are present. Vp and Vs within pillows are quite variable but in general are lower than in massive units, and Vp/Vs is higher. Velocities within breccias are similar to or lower than those of pillows. Within a given morphology the elastic properties do not depend on spreading rate. Within pillows, velocity variations depend primarily on the degree of alteration infilling and the mineralogy of the infilling material, and secondarily on conditions at emplacement. Coherent shear arrivals are sometimes not found within shallow pillows, and in extreme cases the compressional arrival is also incoherent, because of scattering resulting from the similarity between the sonic wavelength and the pillow size. This “sampling bias” in the sonic log is the most likely cause of somewhat higher average sonic than seismic velocities measured within extrusive basalts at shallow depths. Other causes include the fact that laterally finite massive units may be oversampled by borehole measurements and that voids between pillows and fractures, whose size is within an order of magnitude of the sonic wavelength, have less effect on sonic than on seismic velocities. The crustal section can be separated only roughly into seismic layers on the basis of the sonic velocities. Changes in velocity with depth within the extrusive section at a single site giving rise to previously defined seismic layers 2A and 2B are due to a change with depth in the properties of the pillows, associated with differences in the amount and type of infilling materials, rather than to a change with depth in the relative proportion of pillows and massive units. Because alteration occurs early and its history is different at different depths, depth-dependent variations in pillow properties are (1) established at young age and (2) persist even in old crust. Within the intrusive section of DSDP hole 504B, velocities are similar to or higher than those measured in saturated minicores at atmospheric pressure, and seismic and sonic velocities are approximately equal, suggesting that little large-scale porosity is present.

The conditions necessary for compressive and tensile failure of well bores drilled into crystalline rock can be adequately represented by simple elastic failure criteria, and analysis of well bore failure can provide constraints on the magnitudes of in situ stresses if the strength of the rock is known. When applied to several boreholes drilled into continental crust where there is relatively complete knowledge of stress magnitudes, these criteria enable us to predict the depth at which compressive failure of the well bores is observed. In oceanic crust, breakouts have been observed at depths below 700 m below sea floor in Deep Sea Drilling Project (DSDP) hole 504B, drilled into 5.9 Ma crust south of the Costa Rica Rift, and near the bottom of DSDP hole 395A, drilled into 7.3 Ma crust west of the Mid–Atlantic Ridge. In both cases the azimuth of maximum horizontal compressive stress is roughly perpendicular to the ridge axis. As the unconfined compressive strengths of basalt samples from DSDP hole 504B are generally above 200 MPa (Bauer and Handin, 1985), the existence of breakouts in DSDP holes 395A and 504B requires a highly compressional stress state, where Shmin ∼ Sv and SHmax ≥ 100 Mpa at about 500 m subbasement. These results are consistent with the state of stress inferred from compressional (strike-slip and reverse faulting) earthquake focal mechanisms in young oceanic crust. As ridge push forces are relatively small in young oceanic crust, we concur with previous suggestions that the high horizontal compressive stresses result from the thermoelastic effects of a convectively cooled upper crustal layer overlying a conductively cooling lithosphere.

A study of the effects of macroscopic fractures on P and S wave velocities has been conducted in four wells drilled in granitic rock to depths between 0.6 and 1.2 km. The effect of macroscopic fractures is to decrease both Vp and Vs and increase Vp/Vs. In wells with a relatively low density of macroscopic fractures, the in situ velocity is similar to that of saturated core samples under confining pressure in the laboratory, and there is a clear correlation between zones with macroscopic fractures and anomalously low velocities. In wells with numerous macroscopic fractures, the in situ velocity is lower than that of intact samples under pressure, and there is a correlation between the rate at which in situ velocity increases with depth and the rate at which the velocity of laboratory samples increases with pressure. Differences in in situ P wave velocity between wells cannot be explained solely by differences in the degree of macroscopic fracturing, thus emphasizing the importance of composition and microcracks on velocity. In one highly fractured well the in situ P wave velocity is essentially the same for frequencies ranging from 10 Hz to 20 kHz; this suggests that the macrofractures affect velocity similarly over a broad frequency range. Chemical alteration of rock adjacent to macroscopic fractures appears to play an important role in reducing in situ velocities. Synthetic reflection seismograms generated from the velocity logs suggest that fracture zones are one possible source of deep-crustal reflectors observed on seismic reflection profiles.