MR21A-1760
Ultrasonic P and S Wave Velocities in Carbonates From the Arab Formation, Saudi Arabia, and the Western Canada Sedimentary Basin, Canada
The fact that many of the giant hydrocarbon reservoirs, such as Ghawar field in Saudi Arabia and the Grosmont formation in Alberta, are carbonates make such rocks an important research topic. This is particularly true the bulk of laboratory investigations have focused on siliclatic reservoir materials; more work on carbonates is necessary. Anselmetti and Eberli (1993) have shown that the influence of mineral compositions in carbonates is minimal, and cannot be a reason for large changes in velocities. They found that the velocity variation in carbonates is mainly controlled by porosity and pore type. Assefa et al. (2003) showed that the velocities of rocks with high aspect ratio pores are greater than those with low aspect ratio. In this study, we are investigating the effect of pressure, porosity, permeability and saturation in carbonate rock. To date, we have measured the ultrasonic compressional and shear wave velocities for forty carbonate rocks, thirty eight of them are from Arab formation in Saudi Arabia and two from the Western Canad Sedimentary Basin (WCSB). The samples were measured both in dry and water saturated conditions under different confining pressures that varied from 2.5 MPa to 25 MPa for Arab-D samples and from 5 MPa to 70 MPa for WCSB samples. In order to study the effect of microcrack closure, one Arab and one WCSB samples were measured at confining pressure of 40 MPa and 90 MPa respectively. Higher confining pressures were avoided in order to lower the risk of damaging the samples. Some preliminary results from P and S wave velocity measurements on several samples under dry and water saturation conditions are presented. These measured results and the calculated water saturated velocities from Gassmann's equation are compared
MR21A-1761
Experimental Study of the Roles of Mechanical and Hydrologic Properties in the Initiation of Natural Hydraulic Fractures
Natural hydraulic fractures (NHFs) are inferred to form where pore fluid pressure exceeds the least compressive stress; i.e., where the hydraulic fracture criterion is met. Although it has been shown that mechanical heterogeneities serve as nuclei for NHFs, the relative roles of mechanical anisotropy and hydrologic properties in initiating NHFs in porous granular media have not been fully explored. We designed an experimental protocol that produces a pore fluid pressure high enough to exceed the hydraulic fracture criterion, allowing us to initiate NHFs in the laboratory. Initially, cylindrical samples 13 cm long and 5 cm in diameter are saturated, σ1 is radial, and σ3 is axial. By dropping the end load (σ3) and pore fluid pressure simultaneously at the end caps, we produce a large pore fluid pressure gradient parallel to the long axis of the sample. This allows us to meet the hydraulic fracture criterion without separating the sample from its end caps. The time over which the pore fluid remains elevated is a function of hydraulic diffusivity. An initial test with a low diffusivity sandstone produced NHFs parallel to bedding laminae that were optimally oriented for failure. To evaluate the relative importance of mechanical heterogeneities such as bedding versus hydraulic properties, we are currently investigating variably cemented St. Peter sandstone. This quartz arenite exhibits a wide range of primary structures, from well developed bedding laminae to locally massive sandstone. Diagenesis has locally accentuated these structures, causing degree of cementation to vary with bedding, and the sandstone locally exhibits concretions that form elliptical rather than tabular heterogeneities. Bulk permeability varies from k=10-12 m2 to k=10-15 m2 and porosity varies from 5% to 28% in this suite of samples. Variations in a single sample are smaller, with permeability varying no more than an order of magnitude within a single core. Air minipermeameter and tracer tests document this variability at the cm scale. Experiments will be performed with σ3 and the pore pressure gradient both perpendicular and parallel to sub-cm scale bedding. The results of these tests will be compared to those of structurally homogeneous samples and samples with elliptical heterogeneities.
MR21A-1762
Influence Of Lixiviant Flow Rate On Heap Leaching Of Low Grade Manganese Carbonate Ore
A study was carried out to ascertain the influence of lixiviant flow rate on leaching of low grade manganese carbonate ore from Nsuta, Ghana, and also the levels of impurities in the resulting leachate when the ore is heap-leached for 24 hours at two different flow rates. It was found out that the lixiviant flow rate has influence on the dissolution of manganese; about 8.42% recovery of manganese was obtained when the ore was leached at a flow rate of 5 ml/min whilst 99.28% was obtained in the case of 10 ml/min. It was also observed that some levels of impurities such as iron and magnesium were in the leachate: 4075 mg/l of iron and 2575 mg/l of magnesium were in the leachate when the ore was leached at a flow rate of 5 ml/min and 2500 mg/l of iron and 3970 mg/l of magnesium were in the leachate in the case of 10 ml/min. The results indicate that leachate should be purified before the final recovery of manganese from solution.
MR21A-1763
EQUIVALENT ELASTIC MODELS
We pose a question: can a theoretical inclusion model, specifically, the differential effective medium model (DEM), be used to match experimental velocity data in rocks that are not necessarily made of inclusions (such as clastics)? It is indeed possible in some cases by using an almost constant inclusion aspect ratio within wide ranges of porosity and mineralogy. We first approach this question by using empirical velocity- porosity equations as proxies for data. By finding a DEM inclusion aspect ratio (AR) to match these equations, we find that the required range of AR is remarkably narrow. Moreover, a constant AR of about 0.13 can be used to accurately match empirical relations in competent sand, shale, and quartz/calcite mixtures. We show that this discovery can be practically utilized to (a) predict from ; (b) describe velocity- frequency dispersion between low-frequency and ultrasonic experiment; (c) predict the dry-frame elastic properties from ultrasonic data on liquid-saturated samples (where Gassmann's fluid substitution is not applicable); (d) predict the attenuation of P-wave velocity; and (e) establish tight constrains for the ranges of possible variation of and at a given porosity in some mineralogies. We also apply this approach to laboratory data (rather than empirical equations) and confirm a positive answer to the main question with all applications of this result still valid.
MR21A-1764
Experimental Study of Water-Sandstone/Limestone Reaction under Triaxial Stress
Hydrochemical anomalies normally occurred before and after the earthquake, but not all of hydrochemical anomalies were corresponding to earthquake. Genetic mechanisms of hydrochemical anomalies related to earthquake could help to distinguish them. Hydrochemical variations in the saturated sandstone and limestone with creeping and shearing deformations under triaxial compression were experimentally investigated, and genetic mechanisms of hydrochemical anomalies related to earthquake were discussed. The time-dependent and stress-dependent experiments were conducted by loading differential stress to the sealed sample systems. Concentrations of anions and cations in the produced solutions under different stresses and time were determined with IC and ICP-OES, respectively. Anions include F-, Cl-, NO3- and SO42-, and cations include Ca2+, Mg2+, K+ and Na+. The experimental results showed that the ion concentrations increased with the increasing of crept time and the differential stress loaded on the saturated rocks, except that the concentration of Mg2+ in the experiments of water-sandstone reaction decreased. Concentrations of F- and Cl- increased by about 10 and 4 times in the experiments of water-sandstone reaction, and concentration of F-, Cl- and Ca2+ increased by about 7, 6 and 5 times in that of water-limestone reaction, respectively. The results indicated that rocks generated deformation and micro-cracks under differential stress, and this probably resulted in dissolution of minerals, ion-exchange because of the increasing of water-rock reaction area, and releasing of the closed fluid inclusion. Finally positive and negative anomalies of hydrochemistry were brought. Concentrations of F- and Cl- change obviously in the product solutions of sandstone samples, so F- and Cl- may be considered as the sensitive components for monitoring earthquake in the sandstone aquifer, and F-, Cl- and Ca2+ are relatively sensitive in the limestone aquifer. This research was financially supported by the basic research Foundation from Institute of Earthquake Science, China Earthquake Administration (0207690233); the Earthquake Science Foundation, China (C07002) and the National Natural Science Foundation of China (40774036).
MR21A-1765
Nonlinear Stress/Strain Behavior of a Synthetic Porous Medium at Seismic Frequencies
Laboratory experiments on porous core samples have shown that seismic-band (100 Hz or less) mechanical, axial stress/strain cycling of the porous matrix can influence the transport behavior of fluids and suspended particles during steady-state fluid flow through the cores. In conjunction with these stimulated transport experiments, measurements of the applied dynamic axial stress/strain were made to investigate the nonlinear mechanical response of porous media for a poorly explored range of frequencies from 1 to 40 Hz. A unique core-holder apparatus that applies low-frequency mechanical stress/strain to 2.54-cm-diameter porous samples during constant-rate fluid flow was used for these experiments. Applied stress was measured with a load cell in series with the source and porous sample, and the resulting strain was measured with an LVDT attached to the core face. A synthetic porous system consisting of packed 1-mm-diameter glass beads was used to investigate both stress/strain and stimulated mass-transport behavior under idealized conditions. The bead pack was placed in a rubber sleeve and static confining stresses of 2.4 MPa radial and 1.7 MPa axial were applied to the sample. Sinusoidal stress oscillations were applied to the sample at 1 to 40 Hz over a range of RMS stress amplitude from 37 to 275 kPa. Dynamic stress/strain was measured before and after the core was saturated with deionized water. The slope of the linear portion of each stress/strain hysteresis loop was used to estimate Young's modulus as a function of frequency and amplitude for both the dry and wet sample. The modulus was observed to increase after the dry sample was saturated. For both dry and wet cases, the modulus decreased with increasing dynamic RMS stress amplitude at a constant frequency of 23 Hz. At constant RMS stress amplitude, the modulus increased with increasing frequency for the wet sample but remained constant for the dry sample. The observed nonlinear behavior of Young's modulus and the dependence of stress/strain hysteresis on strain amplitude and frequency have implications on how seismic waves can influence the mechanical properties of granular porous materials in the Earth. This work was funded by the U.S. Department of Energy Basic Energy Sciences Program under the Los Alamos National Laboratory contract no. DE-AC52-06NA25396.
MR21A-1766
Spectral Seismic Signatures for Petrophysical Property Estimation of Thin Sand-Shale Reservoirs
Reservoir property estimation using seismic response is common in petroleum exploration. However, detecting and estimating petrophysical properties of thinly layered reservoirs with layers below seismic resolvability can be challenging. In this study, we investigate seismic signatures of thin sand-shale sequences using numerical Monte-Carlo simulations. The thin sand-shale sequences are modeled as a first-order, discrete, 1-D Markov chain. We create a set of transition matrices to simulate aggrading sand-shale sequences. Properties (e.g. density, velocity) assigned to layers in the sequences come from rock-physics relations, particularly those relating sand-shale ratios to porosity and velocity. Synthetic seismograms for multiple realizations of the sequences are then generated and used to extract parameters based on the wavelet transform spectrum of the seismogram. We plot the variance of wavelet-coefficient modulus versus scales on a log-log scale and obtain the slope and intercept of a linear fit to this plot. Our two main scenarios in this study aim to explore the effect of sand fractions and saturation. For the selected form of transition matrices with the same setting of water saturation, an increase in proportions of sand in the total sequences results in a decrease in both slope and intercept values of the wavelet spectrum. For the same transition matrix, an increase in water saturation may result in an increase in both slope and intercept. Results obtained from various transition matrices with a similar proportion of sand do not always show the same behavior in the wavelet slope-intercept space. These differences may arise depending on higher-order spatial statistics of the layered media, for example, how multiple sand layers are distributed in the sequences.
MR21A-1767
Uncertainties in Rock Properties and Effects on Seismic History Matching
Seismic time-lapse data has begun to play an important role in reservoir history matching. Time-lapse 4D seismic data can provide information on the dynamics of fluids in the reservoir based on the relation between variations of seismic signals and movement of hydrocarbons and changes in formation pressure. The calibrated reservoir models, optimally constrained to both integrated flow response and spatially distributed seismic responses, will give a better description of the reservoir, and consequently, more reliable forecasts. A typical step in seismic history matching is to compute the effects of changes in saturation and pore pressure on the seismic velocities. The changes in seismic velocities can be modeled using Gassmann's equations .Parameters that are updated during history matching include porosity, permeability, and lithology or facies. One of the rock parameters that is often taken to be constant in many flow simulators is the rock pore space compressibility with respect to pore pressure. Porous rocks are described by at least four inter-related compressibilities, one of which affects the flow while the other impacts the wave velocities. The rock compressibility affects both the flow, and the seismic wave velocities. This poster explores the sensitivity of flow response and seismic velocity changes to variations in rock pore compressibility. What might be some of the pitfalls in time-lapse modeling that might result from ignoring uncertainty in pore compressibility? Using the Stanford VI synthetic reservoir we show how uncertainties in pore compressibilities can lead to uncertainties in the time-lapse seismic response and thus impact overall history matching.
MR21A-1768
Developing Mechanisms of Braided Fault Patterns and Asymmetric Fault Breccias across Reactivated Fault Zones
This study examines mechanisms that resulted in asymmetric fault breccias patterns in Cretaceous sedimentary rocks of the Gyeongsang supergroup in SE Korea. Exposures of the supergroup include mostly sandstones and shales, some of which are hornfelsic due to later igneous activities. Several reactivated faults are present in the study area, including a 1 km-long, 2 m-thick, N-S striking, east-dipping fault. Grid- mapping, thin section observation of the fault gouge zone, and analysis of lineations were undertaken to understand the structural pattern and deformation history of the fault system. The main N-S striking fault zone shows various fault gouge bands and S-C fabrics that indicate discrete stages of normal and reverse movement during several stages of deformation. The relative timing of kinematic indicators, such as cleavages, lineations, and slickenlines, suggest that the fault underwent early normal displacement during SE extension and was later reactivated during NNW compression, resulting in reverse movement. Most major fault sets do not crosscut each other; instead, the fault gouges within the fault zone split and merge into other fault zones. The braided fault patterns in this fault system indicates it is a mature fault systems. In the study area, breccias and gouges within the brittle fault zones have irregular patterns, which include lens-shaped blocks, and asymmetric distributions of grain size and fracture density across the fault zone. The asymmetric pattern across the fault zone, suggests a complex mechanical model that involved several episodes of reactivation.
MR21A-1769
Estimation of underground fracture density of granitic rocks body using P-wave velocity tomography and crack tensor theory
A crystalline rock such as granitic rocks is one of a candidate for deep underground excavations for disposal
of high-level radioactive waste. Such a rock is fractured medium, and such fractures cause alternation of the
mechanical properties such as strength and deformability, and the hydrological properties such as
permeability and diffusion coefficient since the fractures function as a water-pathway. Therefore, the
fractures are a very important parameter, when we determine hydro-mechanical properties of the crystalline
rock at great depth. When the underground water flow is estimated by numerical simulation, distribution of
the underground permeability is an important initial value to make hydro-geological model. However, it is
difficult to obtain the underground permeability.
Underground permeability is able to represent permeability tensor Pij (Oda, 1987), and Pij is function of the
geometry of cracks such as density, aperture and connectivity, therefore, we are able to estimate
underground permeability using the geometry of cracks. A non-dimensional second rank tensor Fij, called the
crack tensor, has successfully been introduced to deal with crack geometry. We seek the possibility of using
the longitudinal wave velocities to overcome the difficulty associated with the determination of crack tensors.
A new second-rank tensor Vij is introduced by takemura and oda (2005), such that the longitudinal wave
velocity is represented in terms of the tensor, and the crack tensor Fij is then given as a function of Vij. In this
study, we try to estimate two-dimensional underground permeability using the distribution of underground
longitudinal wave velocity, which is tomography image of the longitudinal wave, Vij and permeability tensor
based on crack tensor theory. In addition, the aperture of fractures under confining pressure needs to
estimate permeability tensor, and we propose experimental formula to estimate the aperture of fracture in
underground.
http://www.geo.chs.nihon-
u.ac.jp/geotec
MR21A-1770
Application of X-ray CT Imaging to the Study of Compaction Localization in High Porosity Sandstones
Compaction localization in high porosity sandstones was documented over the last decade both in field studies and laboratory experiments. For the latter, despite an increasing number of experimental data, the parameters controlling the nucleation and propagation of localized compaction features as well as the associated spatial distribution of damage remains however unclear. Part of the difficulty lies in finding an appropriate way of describing the phenomenon since compacted areas may be associated with relatively limited density changes and compacted features may not form as a single object but often appear at multiple locations within the sample. Pilot studies involving X-ray CT imaging suggested that porosity, pore space heterogeneity and composition play a role in the development of compaction localization. We present results obtained from X-ray CT scans of five sandstones acquired at 50 microns resolution in which various degrees of compaction localization (from highly localized to homogeneously distributed) have been observed. While raw X-ray attenuation data do not allow to discriminate between compacted and relatively undamaged zones, it has been shown in previous studies that the coefficient of variation (COV) (standard deviation divided by the mean) calculated over volumes of 3*3*3 voxels could image significant changes in density homogeneity associated with compaction. Since homogenization can be achieved through compaction as well as through a change in spatial resolution, we mapped the distribution of the COV for a stepwise decrease of resolution. For all the imaged sandstones, a volume corresponding to a maximum heterogeneity (higher COV values) was identified. We also drew these maps for images obtained before and after compaction in order to probe the effect of compaction on the COV statistics. These results were then compared with mercury porosimetry data. Our study concludes that X-ray CT data can help predicting and identifying the formation of localized compaction features in sandstone.
MR21A-1771
Confocal Laser Scanning and Atomic-Force Microscopy in Estimation of Elastic Properties of Organic-Rich Rocks, Bazhenov Formation, Russia.
We estimate the indentation modulus (related to Young's modulus via Poisson's ratio) of organic-rich shale samples using a nano-indentation technique based on atomic-force microscopy, coupled with scanning electron microscopy (SEM) and confocal laser scanning microscopy (CLSM). Our approach is based on elaboration of data from two types of microscopy (SEM and CLSM to separate organic-rich (kerogen containing) regions from the mineral matrix of oil shales, with subsequent nanoscale probing via AFM. First, the microtexture of shales is characterized by SEM. Possible regions of interest are selected, and CLSM imaging is performed to confirm the presence of organic matter. Then, an AFM-based nano- indentation probe is employed to test the hardness of the previously identified region. Finally, nano- indentation modulus values are determined for individual mineral and organic-rich phases. This allows mapping of the absolute value of the modulus, providing spatial variation of elasticity, which then can be correlated with the initial mineralogy of the sample.
MR21A-1772
Reflectivity and Transmissivity of a Water-saturated Porous Plate: First Observations of Slow Biot Wave Conversions on Reflection
Three distinct body wave modes, the fast and slow P and an S wave, propagate in a liquid saturated porous
and permeable solid. The slow, or Biot, wave is highly attenuated and while it may be impossible to observe
directly in the earth it does influence the energy budget available and as such influences both the seismic
reflectivity and transmissivity of porous materials. There remain few experimental tests of wave propagation
in such materials but there are still questions with regards to mechanisms of seismic attenuation and
completely untested theories regarding the boundary conditions that control reflection and transmission at
interfaces. To overcome this limitation, a series of pulse transmission and reflection tests were made through
and from, respectively, a water-saturated plate of sintered glass beads. A novel ultrasonic goniometer
system consisting of a large ultrasonic transmitter and a near-point source ultrasonic receiver were specially
constructed for these tests. Despite the attempt to emulate plane wave behaviour, the finite diffraction
effects of even the large aperture transducer influenced all the observations. This necessitated that the
transducer response be fully modelled in order to eliminate misinterpretation, and these concepts were
validated first on plates of simple isotropic glasses. The reflectivity and transmissivity of the plate were
observed for incidence angles ranging from -50 ° to +50 °. The reflection responses were well
modeled using the open pore boundary condition assumption. Additionally, the more complex transmission
responses were successfully modeled but only once the viscoelastic response of the dry frame was included
for the fast P and S waves. Notably, the slow P attenuation did not depend on the frame attenuation. Taken
together, these results provide additional support for the dynamic poro-elastic theory of Biot in high porosity
materials. An added serendipitous bonus of the reflectivity experiments is the first, to our knowledge,
observations of reflected conversions between the fast P, the slow P, and the S wave at the second interface
of the plate.
http://www-
geo.phys.ualberta.ca/~doug
MR21A-1773
A method for measuring the dielectric permittivity of rocks
Ground Penetrating Radar is a widely used geophysical technique for different purposes. Some of them are locating tunnels, locating irregularities in salt, and off course the determination of soil water content. The theory behind the technique is that different formations have different dielectric properties and therefore different velocities and attenuation of the electromagnetic waves. The main cause of the attenuation of the electromagnetic waves is the presence of water and or clay. In this paper we present a technique for measuring the dielectric permittivity of rocks and minerals. In this technique the amplitude and phase of the reflected signal at the discontinuity plane between the probe and the sample are measured from which the dielectric permittivity is calculated. The technique operates at frequencies that range between 10MHz and 3GHz. We will be presenting the technique and provide some new measurements on natural and synthetic samples of clay and salt and their mixtures. The synthetic clay-salt mixtures were fabricated by a cold compressing technique. The dielectric spectroscopy of these samples is shown to be sensitive to clay contamination. The presence of as low as 5 % clay in a salt sample changes its dielectric permittivity.
MR21A-1774
Numerical and Experimental Analysis on the Volumetric Expansion Ratio of Rock Mass
The volumetric expansion ratio of rock mass on the subsidence occurrence area can explain why the depth of the surface subsidence is lower than the height of an opening; it is because the empty space of the gangway is filled with the broken rock. However, until now, the volumetric expansion ratio of rock mass has been calculated by the theoretical equation through numerical models, and it has hardly been used in the studies of the subsidence prediction and the restoring plan. Also, the studies on the surface subsidence that did not consider the volumetric expansion ratio have overestimated the surface subsidence to some degree. In this study, a laboratory test was conducted according to the stacking patterns and the stresses of broken gneiss, limestone, and shale in order to analyze the volumetric expansion ratio. And, unlike the existing continuum analysis, the numerical analysis was conducted by discontinum modeling using UDEC(universal distinct element code) in order to analyze the subsidence mechanism quantitatively.
MR21A-1775
The contribution of Anisotropy of Magnetic Susceptibility (AMS), Anisotropy of P-wave Velocity (APV) and Petrophysic analysis in the strain characterisation at Sheep Mountain Anticline (Wyoming, USA)
Using Magnetic Susceptibility (AMS) and Acoustic Velocity (APV) Anisotropy data, combined with petrophysic
analysis and fracture data, we discuss the relationship between fold development and strain distribution. The
natural laboratory chosen for this study is the Sheep Mountain Anticline (Wyoming, USA). This fold is
asymmetric, basement cored, and was formed during the Laramide orogeny in the Early Tertiary.
AMS
results suggest the existence of three different directions of maximum susceptibility axis: one is parallel to fold
axis, the second one is perpendicular to this axis and the third direction is perpendicular to a small
secondary fold. Those directions are carried by the micro-fractures in the limestone formation and by the
grain form in the grainstone formations. The high consistency of the strain ellipsoid orientation with respect to
the anticline geometry exists.
The AMS and APV results show differences of strain ellipsoid between the
forelimb, the hinge line and the backlimb of the anticline. The AMS fabrics in the forelimb are less evolved
than in the backlimb, in agreement with fracture data which show that there are more sets of fractures in the
backlimb than in the forelimb. The P-wave velocity in the backlimb is higher than in the forelimb. The minimum
principal axis for APV tensor is perpendicular to the bedding, while its maximum principal axis is
perpendicular to the strike. We observe also that there's less anisotropy of P-wave velocity along the hinge.
Thus the backlimb apparently suffered less strain, but recorded a more evolved strain pattern, both in term of
microstructures and of matrix scale layer parallel shortening. We use these results to discuss the kinematics
and fold evolution.